Proper Names As Cultural Referents in British Chick Lit: a Corpus-Based Analysis of Their Translations Into Spanish and Italian

Proper Names as Cultural Referents in British Chick Lit: A corpus-based analysis of their translations into Spanish and Italian

Laura Cantora Tuñón

Submitted in accordance with the requirements for the degree of Doctor of Philosophy

The University of Leeds School of Modern Languages and Cultures

October 2013

The candidate confirms that the work submitted is her own and that appropriate credit has been given where reference has been made to the work of others.

This copy has been supplied on the understanding that it is copyright material and that no quotation from the thesis may be published without proper acknowledgement.

Assertion of moral rights: The right of Laura Cantora Tuñón to be identified as Author of this work has been asserted by her in accordance with the Copyright, Designs and Patents Act 1988.

Acknowledgments

First and most importantly I would like to express my most heartfelt gratitude to my supervisor Professor Jeremy Munday for all his encouragement, support, advice and patience during these years in which he help me grow from a student into a researcher. It has been an honour working with him. I am also grateful to his family for so openly and warmly welcoming me, and my family, into their home and lives.

Professor Anthony Hartley’s input during the early stages of this project is also greatly appreciated. As is the assistance with the technology offered by Dr Bogdan Babych and other colleagues at the Centre for Translation Studies of the University of Leeds. I am most grateful. I would also like to mention Dr Carmen Cuéllar Lázaro, from Valladolid University in Spain, for introducing me to this fascinating world of translated proper names.

The warmest thank you goes to my father, sister and all my family in Spain for always believing in me. To my Godmother, Ana, and my best friend, María José, for being by my side at every step of the way, smiling at all the highs and hugging at all the lows. To my little Arwen, for being the most wonderful thing ever. And lastly, the biggest thank you to my Grant, for all the insightful conversations; because I could not have arrived to this point without his love, belief and commitment.

This Thesis is dedicated to the memory of my Mother

iii

Abstract

This thesis considers the role that proper names play in British mass market women’s fiction (genre generally known as Chick Lit) by means of an analysis of the way in which those proper names have been translated into Spanish and Italian. The relationship between source text and target text is explored in terms of the procedures employed to translate each proper name.

Our main hypothesis is that Toury’s (1995) law of growing standardization would prevail in the translated text in an attempt to bring the connotations and foreign elements of the source language text nearer to the target language readership. This hypothesis is tested by creating a parallel trilingual corpus containing three novels in their original English versions as well as their translations into Spanish and Italian. Using corpus processing tools, such as named-entity recognition programs and trilingual concordancers, a total of 1500 distinct proper names are extracted from these texts which means that 3000 translation instances are available for analysis. These names are categorised using a novel faceted taxonomy, and the translations are classified using a purpose-built map of translation procedures. These categorisations are then analysed using corpus processing measurements such as raw and relative frequency to detect any possible patterns in the translation of proper names in modern literature.

The research shows that translators attempt to identify an ‘equilibrium’ between conserving certain elements and explaining others; while attempting to bring the connotations and foreign elements closer to the target language culture a wide variety of procedures are used. The data also reveals a more marked tendency to conserve the proper names in their original form in the Spanish target texts than in the Italian target texts, where there is more variation in the range of procedures employed.

In addition to these results, a major contribution to knowledge of this research is the creation and successful application of the taxonomy and map of procedures and the processes devised for the analysis which are now available to be used in other investigations.

Keywords: proper name, corpus-based translation studies, translation procedure, norms and laws of translation.

Table of Contents

Acknowledgments ...... iii Abstract ...... iv Table of Contents ...... v List of Figures ...... x List of Tables ...... xi List of Abbreviations ...... xii Introduction ...... 13 1. Introduction ...... 13 2. Hypothesis ...... 14 3. Methodology ...... 15 4. Research Questions ...... 17 5. Aims and Objectives ...... 18 5.1 Concrete Outputs ...... 18 6. Thesis Structure ...... 19 6.1 Arrangement of Chapters ...... 21 7. Contribution to knowledge ...... 26 Chapter 1. Theoretical Framework ...... 28 1.1 Introduction ...... 28 1.2 Holmes’ Map ...... 28 1.3 Toury’s Descriptive Translation Studies methodology and concepts ...... 30 1.4 The quest for Translation Universals ...... 34 1.4.1 Corpus Based Translation Studies (CBTS) ...... 36 1.5 Conclusions ...... 41 Chapter 2. In Search of a Definition of Proper Name ...... 43 2.1 Introduction ...... 43 2.2 Linguistic approach to describing proper names...... 44 2.2 Philosophical approach to the meaning of proper names...... 46 2.4 A definition of Proper Name ...... 55 2.5 Conclusions ...... 56 Chapter 3. Proper Names and Translation ...... 57 3.1 Introduction ...... 57 3.2 Arguments against the translation of proper names ...... 58 3.3 Arguments in favour of the translation of proper names ...... 61 3.4 Conclusions ...... 71 v

Chapter 4. Translation Procedures and Proper Names ...... 73 4.1 Introduction ...... 73 4.2 Vinay and Darbelnet and the rise in empirical research into translation procedures ...... 75 4.3 Translation procedures for culture-bound elements ...... 77 4.3.1 Newmark (1988) ...... 78 4.3.2 Hermans (1988)...... 80 4.3.3 Franco Aixelá (2000) ...... 83 4.3.4 Leppihalme (1997) ...... 86 4.3.5 Marco (2004) ...... 88 4.4 Conclusions ...... 91 Chapter 5. Chick Lit ...... 93 5.1 Introduction ...... 93 5.2 Definition of a new genre ...... 93 5.3 Chick Lit in Spain and Italy ...... 99 5.4 Summary of selected books ...... 102 5.4.1 Introduction ...... 102 5.4.2 Amanda’s Wedding...... 102 5.4.3 My Life on a Plate ...... 104 5.4.4 The Diary of Bridget Jones ...... 105 5.5 Conclusions ...... 106 Chapter 6. Corpus Methodology: Data Acquisition ...... 108 6.1 Introduction ...... 108 6.2 Corpus design, compilation and interrogation ...... 109 6.3 Creating BRIDGET ...... 123 6.3.1 Design ...... 123 6.3.2 Compilation ...... 127 6.3.3 BRIDGET ...... 131 6.4 Interrogating BRIDGET ...... 132 6.4.1 Identifying Proper names in STs ...... 133 6.4.1.1 GATE ...... 136 6.4.1.2 Word-frequency lists ...... 138 6.4.1.3 Access for BRIDGET ...... 142 6.4.2 Identifying Proper Names in TTs ...... 148 6.4.2.1 Concordancing ...... 148 6.5 Conclusions ...... 150

Chapter 7. Data Analysis 1: Classifying the Proper Names...... 152 7.1 Introduction ...... 152 7.2 Taxonomy for Proper Names ...... 155 7.2.1 Ontological Facet ...... 155 7.2.1.1 Introduction – The ontological classification of the Prolex Group ...... 155 7.2.1.2 BRIDGET’s ontological facet ...... 160 7.2.2 Semantic Facet ...... 183 7.2.3 Scope Facet ...... 187 7.2.4 Sphere Facet ...... 189 7.3 Methodology ...... 190 7.4 Conclusions ...... 197 Chapter 8. Data Analysis 2: Classifying the Translation Procedures ...... 199 8.1 Introduction ...... 199 8.2 Proposal of Map of Procedures ...... 202 8.2.1 Relation to previous maps ...... 203 8.2.3 Fixed Translation ...... 208 8.2.4 Transcription ...... 209 8.2.5 Literal Translation ...... 209 8.2.6 Gloss ...... 210 8.2.7 Explicitation ...... 212 8.2.8 Adaptation ...... 215 8.2.9 Generalization ...... 217 8.2.10 Cultural Replacement...... 218 8.2.11 Cultural Substitution ...... 218 8.2.12 Omission ...... 220 8.2.13 Insertion ...... 221 8.2.14 Compression (Marco), Label and Couplet (Newmark) ...... 221 8.3 Continuum ...... 223 8.4 Methodology ...... 229 8.5 Conclusions ...... 231 Chapter 9. Results and Findings ...... 232 9.1 Introduction ...... 232 9.2 Translation procedure usage - raw and relative frequency ...... 235 9.2.1 Introduction ...... 235 9.2.2 Overview ...... 237

vii

9.2.2.1 – Lesser used procedures ...... 240 9.2.2.2 Conclusions ...... 250 9.2.3 Book 1 ...... 251 9.2.4 Book 2 ...... 255 9.2.5 Book 3 ...... 260 9.2.6 Overall summary ...... 265 9.2.6.1 Spanish ...... 265 9.2.6.2 Italian ...... 268 9.2.7 Conclusions ...... 272 9.3 Procedures & Taxonomy ...... 273 9.3.1 Introduction ...... 273 9.3.2 Ontological facet ...... 274 9.3.3 Scope Facet ...... 279 9.3.4 Sphere Facet ...... 280 9.3.5 Semantic facet ...... 282 9.3.6 Conclusions ...... 285 9.4 Names in context – qualitative analysis ...... 286 9.4.1 Introduction ...... 286 9.4.2 Anthroponyms ...... 288 9.4.3 Toponyms ...... 302 9.4.4 Chrematonyms ...... 308 9.4.5 Pragnonyms ...... 312 9.4.7 Ergonyms ...... 315 9.4.8 Conclusions ...... 320 9.5 Conclusions ...... 323 Chapter 10. Discussion of the Results ...... 328 10.1 Introduction ...... 328 10.2 Exploring the relationship of the findings to TS ...... 328 10.3 Conclusions ...... 341 Chapter 11. Conclusions ...... 343 11.1 Introduction ...... 343 11.2 Summary of the work undertaken ...... 344 11.3 Aims and objectives ...... 345 11.4 Research questions ...... 353 11.5 Limitations and further work...... 360

viii

11.6 Final remarks ...... 363 Bibliography ...... 364 Primary texts ...... 364 Secondary sources ...... 365

List of Figures

F. 1.1 Holmes’ conception of Translation Studies 30

F. 4.2 Marco’s Continuum 89

F. 6.3 Sample of concordances for unexpected from TEC 122

F. 6.4 Example of results from a search of ‘Bridget Jones’ in the Spanish 125

database of books edited in Spain.

F. 6.5 Excerpt from the purpose-built script for GATE 137

F. 6.6 Excerpt from frequency list for Amanda’s Wedding 139

F. 6.7 Example of the table used to modify results from “Access for BRIDGET” 146

F. 6.8 Sample of concordances for “amanda” from the multilingual concordancer 149

F. 7.9 Excerpt from Amanda’s Wedding spreadsheet for classifying SL names 192

F. 7.10 Excerpt from Amanda’s Wedding spreadsheet for classifying TL names: SP 192

F. 7.11 Excerpt from Amanda’s Wedding spreadsheet for classifying TL names: IT 192

F. 8.12 Procedures plotted on a continuum 226

F. 8.13 Excerpt from Amanda’s Wedding spreadsheet for classifying the 230 procedures

F. 9.14 Procedure relative frequency against continuum – Amanda’s Wedding 254

F. 9.15 Procedure relative frequency against continuum – My Life on a Plate 259

F. 9.16 Procedure relative frequency against continuum – Bridget Jones’s Diary 264

F. 9.17 Relative frequency for procedures in Spanish TTs 266

F. 9.18 Procedure relative frequency in Spanish TTs against continuum 268

F. 9.19 Relative frequency for procedures in Italian TTs 269

F. 9.20 Excerpt from My Life on a Plate’s glossary in Italian TT 271

F. 9.21 Procedure relative frequency in Italian TTs against continuum 272

F. 10.22 Continuum set against Toury’s initial norm 330

List of Tables

Table 6.1 Representation of Laviosa’s corpus typology 111

Table 6.2 Semi-automatic alignment: replacement 129

Table 6.3 Outputs of the different retrieval methods for Amanda’s Wedding 148

Table 7.4a Classification of Proper Names by Prolex Group 158

Table 7.4b Prolex classification highlighting differences with Bridget’s taxonomy 182

Table 7.5 Anthroponym hypertype with its types 163

Table 7.6 Toponym hypertype with its types 167

Table 7.7 Ergonym hypertype with its types 170

Table 7.8 Pragnonym hypertype with its types 172

Table 7.9 Chrematonym hypertype with its types 174

Table 7.10 Ontological facet with its hypertypes and types 176

Table 8.11 Map of procedures with examples from BRIDGET 203

Table 8.12 Relationship between the procedures in previous models and map 205 used with BRIDGET

Table 8.13 Representation of the changes to the referents 227

Table 9.14 Books in the corpus with number of translation instances analysed 233

Table 9.15 Procedure relative frequency across all TTs 238

Table 9.16 Breakdown of procedures used in each TT – Amanda’s Wedding 251

Table 9.17 Breakdown of procedures used in each TT – My Life on a Plate 255

Table 9.18 Breakdown of procedures used in each TT – Bridget Jones’s Diary 260

List of Abbreviations

BNC – British National Corpus

CBTS – Corpus Based Translation Studies

DTS – Descriptive Translation Studies

EN – English

F - Figure

IT – Italian

P - Page

PNs – Proper Names

SL – Source Language

SP – Spanish

ST – Source Text

TL – Target Language

TS – Translation Studies

TT – Target Text

xii

Introduction

“I have as many names as there are winds, as many titles as there are ways to die” Gaiman, N.: American Gods (p.145)

1. Introduction

The study of proper names has occupied linguists, grammarians, logicians and philosophers since Aristotle. Names are carefully selected to play a special role in fictional works and to evoke certain connotations in the reader. Recreating this effect has always been one of the biggest challenges that a literary translator faces (Nord

2003, Cuéllar Lázaro 2000, Zabeeh 1968). This research will study this problem by means of an analysis of proper names extracted from mass market women’s fiction, a genre commonly known as Chick Lit, and their translations into Spanish and

Italian.

The Chick Lit genre has specifically been chosen as proper names play a heightened role in these novels in shaping the scene and as such they open a window on the socio-cultural reality of the source culture. These novels are set in contemporary society and make use of real proper names both to describe scenes and to evoke certain social connotations for the reader (Cantora 2010). Thus, if the character plans a night out in London she will be planning to go to real locations in modern

London (such as the Princess Louise pub in Holborn or to a dinner party at The Ivy); if she watches television she would watch real programmes that are currently showing (EastEnders, Football Focus, Holby City), and so on. This feature makes

13 these novels very rich in proper names1 and more specifically in culturally embedded proper names which open up a big gap between source text (ST) and target text

(TT) reader and represent a considerable challenge for the translator. The focus of this research will fall on the procedures and methods employed by translators when it comes to transmitting the connotations, perceptible for native readers, into a new culture, with different traditions, perceptions and values.

2. Hypothesis

Hermans (1988) suggests that through the analysis of translated proper names it should be possible to infer the different translational norms at play in any given translation. The current research builds upon this brief work by Hermans and aims to unveil the patterns of the translation of proper names in Chick Lit through an analysis of the translation procedures most commonly employed to transfer them to different languages. It draws on Toury’s (1995:36-9) three-phase methodology for systematic Descriptive Translation Studies (DTS). Selecting a pair of ST and TT segments, comparing the function in the original context and the function in the target context and drawing implications from the findings. The trilingual character of the project (English STs and Spanish and Italian TTs) allows for generalizations to be drawn about the behaviour of translators with regards to proper names both from an intra- and inter-linguistic point of view, taking into account what happens in each target language individually as well as comparing both target languages against each other. These results should ultimately model Toury’s initial norm (1995:56-9), setting

1 The three novels analysed in this Project contain over 1,500 distinct proper names. 14 the translator’s behaviour towards adequacy or acceptability. It should also be possible to relate the results to Toury’s laws of translation (1995:267-79). From the standpoint of proper names being cultural elements embedded in the SL culture, it can be hypothesised that the law of growing standardization would prevail in the TT; that is, an attempt to bring the connotations and foreign elements of the SL text nearer to the TL readership would prevail. The approach chosen allows for this hypothesis to be systematically texted for the TLs of this study – Spanish and Italian.

3. Methodology

The investigation will be undertaken using tools from corpus linguistics, which facilitate the study of larger amounts of data than it would be feasible to study manually and therefore makes the analysis and results more comprehensive than would otherwise be possible. The main reasons for using corpus tools in this project are twofold. Primarily the intention is to automate the identification of the proper names in the English texts as well as of their corresponding translations in the

Spanish and Italian versions. Secondly, having an electronic corpus of the novels facilitates the quick and reliable access to the data, and its linguistic context, and therefore turns the analysis more efficient, making it possible to work with more data than it would otherwise be feasible to manipulate manually. In addition, computerised formulas and other corpus processing measurements (such as COUNT

IF and recall and precision) are used in the analysis of the data, ensuring a less error-prone account of translation patterns and facilitating a more accurate interpretation of the results and findings.

More specifically, this research will see the creation of a parallel corpus of three

British Chick Lit novels and their translations into Spanish and Italian. This corpus has been called BRIDGET, in honour of the most popular character in the genre2.

These three novels contain more than 1,500 distinct proper names between them.

Taking into account the trilingual character of this project, this means that BRIDGET offers over 3,000 translation instances for the analysis:

Book Author Title Names in ST Book 1 Jenny Colgan Amanda’s Wedding 420 distinct PNs Book 2 India Knight My Life on a Plate 564 distinct PNs Book 3 Helen Fielding Bridget Jones’s Diary 626 distinct PNs

Total number of translation procedures analysed 3,220 instances

Once all the proper names have been identified, I will establish a classification for proper names and will produce a map of translation procedures for proper names.

Each proper name in the corpus will be allocated to one category in the classification and each translation instance will be allocated to a procedure from the map. The frequency of use of each procedure will be plotted on a continuum, where the extremes model Toury’s concepts of acceptability or adequacy. It should thus be possible to identify the trends in the translation of proper names and to express generalisations as the behaviour of translators with regard to the different types of names. The final outcome should describe the processes which lead from source to

2 It is generally agreed that the genre started with the publication of Helen Fielding’s best-seller The Diary of Bridget Jones in 1996 (Ferris and Young, 2006, Gill and Herdieckerhoff, 2006, Gormley, 2009, Modleski, 2008, Yardley, 2006). 16 target name, helping to illustrate translator behaviour towards proper names and validating or negating the hypothesis formulated.

The findings have a potential application to the classroom in teaching trainee translators the different range of options available to and used by translators for culture-bound elements (Pearson, 2003). The empirical evidence obtained through the use of statistical measurements could be used by researchers working on machine translation to optimize the outcome of the automatic translation of proper names.

4. Research Questions

From this problem statement and with the hypothesis in mind, the main research questions for this project are:

1. to what extent can the cultural background and sense of place evoked

for native readers by a proper name be translated into different target

cultures, with different perceptions, traditions and values?

2. how important are external factors such as the context and the

situation (Halliday & Hasan, 1991) when it comes to understanding the

connotations conveyed by a proper name and to what extent is it possible to

map different information through a classification of proper names and/or of

translation procedures?

3. to what extent is it possible to infer translation procedures and

understand the processes which lead from source to target texts through the

analysis of translated proper names?

4. what can a corpus-based approach offer to the study of translated

proper names and to what extent is it beneficial?

5. Aims and Objectives

Broadly speaking, this project aims to unveil the pattern of the translation of proper names in Chick Lit novels, setting translator behaviour towards either adequacy or acceptability, according to Toury’s initial norm (1995:56-9). In order to attain the aims of this study, the specific objectives that need to be achieved are as follows:

5.1 Concrete Outputs

1. To construct a theoretical framework and elaborate a definition

for the term ‘proper name’. This definition would link the most

salient theories of meaning to the theories of the translation of

proper names and could be appropriately used for the investigation

of the treatment of proper names in the corpus;

2. To create a parallel corpus of Chick Lit novels to be used as the

main source of the data for the investigation;

3. To automate the manipulation of the corpus in order to achieve

maximum efficiency and accuracy through an automatic extraction

of the proper names in SL texts as well as their translations in the

TL texts;

4. To design and practically apply a taxonomy for proper names to

classify and structure the data extracted from the corpus;

5. To design and apply a map of translation procedures for proper

names to pinpoint the techniques employed to transfer proper

names from SL texts to TL texts;

6. To generate and interpret data on the frequency of use of each

procedure alone and in combination with the categories in the

taxonomy in order to explore the tendencies behind the use of each

procedure/name type;

7. To infer implications from the findings and draw generalizations

from them.

6. Thesis Structure

This project opens up with a critical discussion of the main translation theories where the research is grounded (DTS, corpus-based translation studies, proper names theories as well as theories on the translation of proper names) and follows with the development of a model of analysis to be used to explore the data using

19 corpus processing tools. The aim is to produce findings that uncover the patterns in translation procedures and can inform other DTS studies.

The theoretical approach encompasses broad theories and key concepts of translation studies as well as more specific theories on the translation of proper names. It also covers philosophical approaches to the definition of proper names

(referential and semantic theories of meaning) and theories from literary criticism for establishing the main traits of Chick Lit, the genre under investigation.

On the other hand, the practical application section will describe the different stages in the project. Following methodologies and processes used in similar corpus-based projects (for instance, Kenny 2001 or Bosseaux 2007) first of all, the corpus will be created and interrogated. Once all the names are automatically located and extracted from the texts using corpus tools, these names can be classified according to their different traits (e.g. type of name, level of connotations, place in the real world outside the context of the novel, for instance). This broadly corresponds to the textual analysis phase in Toury’s methodology (1995:36-9). The comparison of SL and TL segments is undertaken through the exploration of the translation procedures employed for the translations of the names. This methodology has been widely used in other DTS projects, for instance García Marín (2010) successfully followed a similar method for analysing legal cultural referents in a parallel corpus of legal fiction. A quantitative analysis of the frequency of use of each procedure together with the frequency of use of each procedure within each category in the taxonomy sheds light on the behaviour of the different translators. Implications can

20 then be drawn from these findings, both general and specific, both inter and intra- linguistic. On the evidence supplied, the hypothesis formulated above can be tested, and the tendency of translators towards either adequacy or acceptability can be unveiled. This tendency will form a description of the processes of translation - that is how translators get from SL segment to TL segment. This material can then be woven into the complex paradigm of the general theory of translation envisaged by

Toury (1995).

6.1 Arrangement of Chapters

Chapter 1 locates this study within the Translation Studies discipline and links it to the main theories of the field; these theories will be used to express the results and to frame the findings. Holmes’ (1988) map for translation studies as well as Toury’s

(1995) norms and laws are essential concepts for this study, as is the corpus-based approach introduced by Baker (1993) (also Laviosa, 2002 Olohan, 2004) which provides the tools employed to interrogate BRIDGET. The Chapter concludes by drawing the framework for this project to operate in.

Chapter 2 sets out to offer a definition of proper names. This definition is based on the theories of meaning posed along the years by different language philosophers

(Frege, 1892, Searle, 1967, Russell, 1905, Kripke, 1970/1980) but it also incorporates linguistic criteria (Jespersen, 1924/1964, Algeo, 1973). Comparing and reviewing the different approaches taken to define proper names the Chapter

21 explores the most salient theories to conclude with a set of semantic criteria defining the term.

Chapter 3 links these theories of meaning to the more prominent theories on the translation of proper names, key in the Translation Studies discipline. Scholars have generally taken two opposed views, arguing either for the possibility or impossibility of the translation of proper names (for instance, Newmark (1981) declared proper names ‘outside language’ (ibid.:70) whereas Martinet (1982:398) proposed a method for translating literary proper names and scholars like Moya (2000) and

Franco Aixelá (2000) considered what translators really do when translating proper names). A common ground is identified and the Chapter concludes with the idea that, whatever the position taken, scholars always agree on the necessity of transferring semantically ‘loaded proper names’ (Hermans, 1988).

Chapter 4 follows Newmark (1988) in considering that these semantically loaded proper names are potential targets for the study of translation procedures for culture-bound elements. The chapter starts broadly with an introduction into the study of translation procedures along the history of the discipline, then narrows into looking specifically at scholars who have, in one way or another, approached the study of culture-bound elements from a procedural or cultural point of view: the works of Newmark (1988) Hermans (1988), Franco Aixelá (2000), Leppihalme

(1997) and Marco (2004) are all described and reviewed. These models form the basis of the map of procedures to be used within BRIDGET.

Chapter 5 is devoted to the Chick Lit genre. By describing the traits of the genre and reviewing the more prominent works on the subject to date, the theoretical framework of this project is concluded. The chapter also includes a summary of the fictional works selected for inclusion in the corpus. The selection process for these novels is described in the following chapter as part of the account of the corpus creation and manipulation work undertaken.

Chapter 6 initiates the exploration of the methodology and model of analysis employed. Starting from the definition of corpus and the categorisation of the different types of corpora available for CBTS research, it moves to describing some of the more salient projects in the field with special interest in those which more directly relate to the methodology, tools or processes used with BRIDGET (Kenny,

2001, Bosseaux, 2007, García Marín, 2011). The steps taken in the design, creation and manipulation of BRIDGET are then described with great detail. The Chapter concludes when all the data, that is all the names in the corpus (both SL names and their corresponding translations in the TLs), have been acquired.

Chapter 7 describes the first step in the data analysis: the classification of the proper names. This classification is undertaken according to a faceted classification devised following the latest developments in information science and is based on the theories of meaning described in Chapter 2. The chapter includes a detailed description of each category in the taxonomy as well as an account of the methodology employed to apply this taxonomy to the names from BRIDGET.

Chapter 8 revisits the maps introduced in Chapter 4 (Newmark, Hermans, Franco

Aixelá, Leppihalme and Marco) but from a more critical point of view, with the aim of combining their strong points into one comprehensive map of translation procedures. This map is used to classify all the translation instances of the names in

BRIDGET.

Chapter 9 brings together all the results from the analysis of all the data. Through the use of raw and relative frequencies and similar measurements to account for the use of each procedure, individually as well as in combination with the different categories in the taxonomy, it is possible to model the translator behaviour with regards to proper names. The results are presented both by TL and overall for the corpus as a whole. The Chapter concludes with a set of findings (most prominently the high use of borrowing over any other procedure in the map and the more pronounced manipulation being clearly visible occurring with connotative names) which can be inferred from the interpretation of these results.

Chapter 10 offers a discussion of these findings. Linking them with the theories of translation, meaning and translation of proper names set out on the theoretical chapters.

The Conclusions chapter summarises all the work undertaken and concretely indicates where each individual objective has been met. It also offers answers to the research questions and sets the road to future research in the area. The limitations of the project as a whole are also included in this chapter.

Finally, the bibliography and appendixes complete this thesis.

The appendixes are included in an enclosed CD as they are too large to be printed in a legible format. There are 5 appendixes:

Appendix 1 reproduces the list of Chick Lit novels published in the UK during the

1990s and early 2000s which have been translated to both target languages of this study: Spanish and Italian. Bibliographic information for STs and TTs is given.

Appendix 2 comprises an Excel document with all the data analysis and figures for

Amanda’s Wedding.

Appendix 3 comprises an Excel document with all the data analysis and figures for

My Life on a Plate.

Appendix 4 comprises and Excel document with all the data analysis and figures for

Bridget Jones’s Diary.

Appendix 5 includes a copy of ‘Access for BRIDGET’ – program created to automate the extraction of PNs from the corpus.

The Excel documents contained in appendixes 2, 3 & 4 are made up of several tabs each one containing a different stage in the analysis. The classification of all the names is included in the tab called “class” and the figures (both total and per language) are included in the following four tabs. The classification of the translation

25 procedures is included in the tab called “Procedures” and the following tab includes figures showing the usage of each procedure. Lastly, the tab called “procedures and taxonomy” combines the data from the “class” tab with the data from the

“procedures” tab. In addition, Appendix 2 also has three more tabs, two which correspond to the experiments manipulating BRIDGET (Chapter 6) and the other one, called “inter-annotator agreement”, includes data from the experiment conducted in order to assess the reliability of the taxonomy for proper names

(Chapter 7).

7. Contribution to knowledge

The concrete outputs of this research, the corpus of Chick Lit novels as well as the taxonomy for classifying proper names and the map of translation procedures represent the most solid legacy of this project. These creations will be available for further research and investigations.

In addition, more specifically the experiments conducted with the retrieval systems and the detailed description of the corpus creation can be used by other researchers working in the field. The program created for automatically identifying and extracting proper names is also a contribution from this project.

Another important contribution and the aspect with more potential for generalizability of this research are the methods devised for the analysis of all the data. The combination of name category and translation procedures has never been

26 used before to explore the subject of the translation of proper names. And the processes employed to produce and explore this data, using retrieval systems as well as MS Office Excel and Access, represent a novelty in this kind of analysis as well as a theoretical advance, now available for replication in other investigations.

Lastly, this project provides empirical evidence to pin down the behaviour of proper names in translation; the concrete results obtained from the research contribute to furthening the case studies in DTS and corpus-based translation studies.

Chapter 1. Theoretical Framework

“Only the right name gives beings and things their reality”, she said, “A wrong name makes everything unreal” Ende, M.: The Neverending Story (p. 159)

1.1 Introduction

This chapter reviews Holmes’ (1988/2000) framework for the Translation Studies discipline as well as Toury’s (1995) methodology for Descriptive Translation Studies

(DTS) together with his key notions of norms and laws of translation. The concept of translation universals (Chesterman, 2010, Baker, 1993) is explored in line with the rise of Corpus-Based Translation Studies (CBTS). These concepts form the basis for this research on translated proper names. Holmes’ map allows us to locate this study within the DTS framework and Toury’s three way methodology constitutes our main approach to the data. Toury’s norms facilitate the description of the tendencies shown by the translators of the texts in our corpus and the generalizations that can be extracted from the analysis of the results contribute towards a better understanding of the laws of translation. The main source of the data is a parallel corpus compiled and interrogated under a corpus-based approach to Translation

Studies as introduced by Baker.

1.2 Holmes’ Map

The discipline of Translation Studies emerged from Holmes’ (1972) attempts to bring together the different approaches used to study translated texts under one

28 framework. His seminal paper, ‘The name and nature of Translation Studies’ was not published until 1988 and it was subsequently made widely available when included in Venuti’s Translation Studies Reader (Venuti, 2000/2004)3. For proposing the name as well as for delineating its scope, Holmes is considered the father of Translation

Studies as a discipline.

Within this framework (figure 1.1 below), Translation Studies is divided between the

Pure and the Applied branches. The Pure branch has again two branches, theoretical and descriptive. DTS examine the phenomena of translation whilst the theoretical branch aims to draw a general theory of translation. Results from descriptive projects should feed into the general theory of the theoretical branch. Within each branch there are different sections, depending on the focus or type of project.

Within DTS we find product-oriented, studying existing translations, process- oriented, studying the mechanisms in the mind of the translator that lead to the TT and function-oriented which comprises studies of the implications, functions and role of the different translation in the context of the TT language. On the other hand, within the theoretical branch, the division encompasses different sections for aspects such as different languages, text-types, translation type or time frames, for instance.

The Applied branches comprise all practical matters related to translator training, translation criticism and translation tools and aids. DTS have clear applications to the applied branches, as the results from the descriptive studies of translations feeds directly into translator training as well as into the development of tools (as described

3 The paper has been removed from the 3rd edition of The Translation Studies Reader (Venuti, 2012) 29 in the Introduction when setting out the aims and applications of this particular study on proper names).

Figure 1.1 Holmes’ conception of Translation Studies (Toury 1995:10 from Munday 2012:16)

1.3 Toury’s Descriptive Translation Studies methodology and concepts

This framework was mapped (figure 1.1) by Toury in 1995. In the influential

Descriptive Translation Studies and Beyond (19954), Toury sets forth a methodology for Translation Studies to operate in as well as the basis that he considered necessary for the discipline to grow as a science. Toury was successful at moving

Translation Studies beyond the prescriptive approaches of the 1950s and 60s towards a more systematic approach based on describing real texts and real translations, in an attempt to offer empirical validity to the discipline. This attempt was taken even further with the development of big corpora and the expansion of

Corpus-based Translation Studies, which contributed to the progress in the discipline. DTS benefits from working with large amounts of texts and corpus tools make the mechanical investigation of large quantities of data feasible.

4 Second edition published in 2012 30

On the other hand, Toury’s approach was criticised primarily for focusing too much on the TT (Hermans, 1995:218) as well as for working almost exclusively with literary texts. However, despite these objections, the development of a methodology for working with real texts and real translated texts was an important step forward in the discipline and one that managed to put translation at the centre of empirical research projects and studies. Besides, Toury’s methodology was subsequently successfully also applied to the study of scientific texts or to alternative modes such as, for instance, audiovisual translation (Karamitroglu, 2000).

Descriptive Translation Studies and Beyond sets in motion a three way methodology for studying SL texts and their translated counterparts (Toury, 1995:36-9). This methodology encompassed three phases:

1) Situate the text within the target culture system 2) Undertake a textual analysis of the ST and the TT in order to identify relationships between corresponding segments in the two texts. 3) Attempt generalizations about the patterns identified in the two texts. (ibid.:36-9, in Munday 2012:170)

Toury envisaged that, through the use of this methodology, the ‘attempted generalizations’ formulated in research projects at smaller scales will all come together to form a general theory of translation, and that as more data and texts become the centre of new studies these ‘attempted generalizations’ would all contribute towards describing the artefacts and processes of translation.

Thus, another key concept in Toury’s work is the concept of norms, which he defines as:

The translation of general values or ideas shared by a community – as to what is right or wrong, adequate or inadequate – into performance instructions appropriate for and applicable to particular situations. (Toury, 1995:55)

For Toury the ultimate aim of DTS is to reconstruct the norms in operation during the translation process.

Toury (1999:14-25) sees translation as an activity based and dependant on the culture of the specific society and time period where it originates, and therefore the norms established and accepted in that culture govern the activity of translation. He

(1995:56-9) identifies three kinds of norms in translation which may operate in any given translation: initial norm, preliminary norm and operational norm:

- Initial norm relates to the choice made by the translator with regards to the orientation of the text. If the translator decides to follow the norms of the SL culture then the translation is deemed adequate whereas if, on the other hand, the decision is taken to follow the norms of the TL culture the resulting translation is acceptable.

- Preliminary norms relate to translation policy and directness of translation.

Translation policy refers to the factors involving the selection of texts to be

32 introduced in a culture through translation. Directness of translation relates to cases of translation occurring through an intermediate language.

- Operational norms relate to the decisions taken during the translation process.

These are divided between matricial norms and textual-linguistic norms. Matricial norms relate to the macrostructure of the text (omission or relocation of passages, textual segmentation, etc.). Textual-linguistic norms relate to the microstructure of the text (lexical items, phrases, etc.).

The study of these norms through DTS will ultimately contribute to a better understanding of the laws of translation ‘of how translators translate’ (Pym,

2008:311). As a starting point Toury proposes two laws (Toury, 1995:267-79): law of growing standardization and law of interference.

- The law of growing standardization prevails when the translator selects for the TT linguistic patterns common in the TL, disregarding the forms of the original ST.

- The law of interference, on the other hand, sees ST interference in the TT, that is, linguistic features of the SL are introduced in the TL. This is considered negative if they create non-normal TT patters or positive if the features introduced in the TL are not unusual or do not divert from the common patterns generally used in the TL.

These laws were ‘intended to be descriptive generalizations’ (Chesterman, 2010:40) which ‘may be used to aid future translators and researchers’ (Munday, 2012:189)

33 and have had, together with Toury’s methodology, an extremely wide impact on the way research into translation was conducted.

Nevertheless, Toury’s theories were criticized (most strongly by Hermans (1999),

Venuti (1997) and Tymoczko (1998)) primarily on the grounds of his claims of objectivity of this research methodology and the interpretations extracted from it. As

Venuti sees it the ‘difficulty in separating fact from value in humanistic interpretation’

(Venuti, 1997:363) cannot be disregarded. However, up to what point this is not always the case with any kind of research? Thus, notwithstanding these criticisms, the value of Toury’s methodology, as a conduit to facilitate the comparison of SL-TL segments, lies in its simplicity and ease with which it can be replicated and his concepts of norms and laws of translation have become without a doubt ‘the staples of the descriptive branch of Translation Studies’ (Kenny, 2001:54).

1.4 The quest for Translation Universals

Whereas Toury’s focus was on identifying the mechanisms of translation, that is, the way in which translators get from ST to TT, Mona Baker (1993) introduced the idea of looking for the specific features of translated language, features labelled as

‘universals of translation’ (ibid.). At the same time, Baker proposed the adoption of a methodology derived from corpus linguistics as the best way to identifying these universals. Baker highlighted the benefits from such an approach in a pioneering paper (1993) which is often seen as the stepping-stone in the Corpus-Based

34 approach in Translation Studies (term coined by Laviosa (1998) in a special issue of

Meta dedicated to the subject).

Baker (1993) affirms that these universals of translation can be best visualized through a comparison of translated texts with non-translated texts, and furthermore suggests explicitation, grammatical standardization and increased frequency of common words as recurrent features of translated languages. She points towards the use of corpus tools for testing and validating these hypothesis.

For his part, Chesterman (2004) explores universals looking for a ‘conceptual clarification’ (Chesterman, 2010:40) in an attempt to bring together the concepts of universals and Toury’s laws of translation. Chesterman (2004) proposes two types of universals:

- S- Universals: generalizations about a difference between translations and STs

- T- Universals: generalizations about typical differences between translations and non-translations in TL.

He later on (2010) goes on to stress the benefits that the study of universals has had on the discipline as a whole encouraging ‘translation scholars to develop better empirical research designs, to be more precise about their claims, to work with large corpora, to think more in terms of generating and testing hypotheses’ (ibid.:45).

These advances go hand in hand and, to a certain extent, have been possible due to the adoption in the discipline of corpus linguistics tools which, as Baker (1993)

35 predicted, have facilitated the study of larger amounts of data and have made it possible to look into specific features that can be relatively easily identified and compared via specialized software.

1.4.1 Corpus Based Translation Studies (CBTS)

Corpus linguistics is the discipline that studies linguistic phenomena using examples of real texts from the discourse of a given community. It developed during the late

1950s and 1960s alongside the realization of linguists such as Firth (1968), Sinclair

(1966) and Halliday (1966) of the importance of considering the words in their context rather than in isolation. The widespread availability of computers from the

1980s together with the development of text-analysis software contributed to the progress of the discipline as it made it possible to study large amounts of data as well as allowing for objective verification of the results, giving empirical strength to the linguistic research.

Nowadays, it is almost a default that a corpus would be held electronically. John

Sinclair (1966) was one of the first linguists in the UK to apply computer programs to the study of texts, showing the importance of collocations ‘to take account of how an item predicts the occurrence of others and is predicted by others’ (Anderman and Rogers, 2008:10).

Since the 1980s large corpora have served as a resource for education, linguistic and lexicographic research. Projects like the Collins–Sinclair Cobuild use information derived from corpus analysis to create English dictionary entries and grammars. This

36 project, located at Birmingham University, uses the Bank of English, a corpus that contains more than 500 million words. Gill Francis explains how the corpus is their main source of information: ‘in our approach, the data comes first. The corpus is the major informant, providing the raw information we need in order to describe the language’ (Francis, 1993:139).

As mentioned above, it was Mona Baker’s insight to first point out the implications and extended possibilities of incorporating corpus tools into translation research.

Mona Baker predicted in 1993 a ‘turning point’ in Translation Studies as ‘a consequence of access to large corpora of both original and translated texts’. She ventured that ‘large corpora will provide theorists of translation with a unique opportunity to observe the object of their study and to explore what it is that makes it different’ and that ‘it will also allow us to explore, on a larger scale than was ever possible before, the principles that govern translational behaviour’ (Baker,

1993:235). Twenty years later it can be affirmed that her predictions have become true, and now more than ever, corpus-based translation studies are proving both useful and fruitful. Large corpora of translated text, such as TEC (Translational

English Corpus) at Manchester University, offer an important source of information for translators and translation scholars and strengthen the discipline. At the same time other individual projects at smaller levels, like the ones described in Chapter 6, demonstrate that through the analysis and comparison of translated texts (either with their STs or with other translated texts) using corpus tools, it is possible to understand ‘at a much bigger scale than ever before’ (ibid.), the processes that lead to the final TT as well as to identify recurrent features of translated language.

Undoubtedly, many changes have taken place between those first approaches into corpus and translation and the developments of today. Laviosa (2011) presents a compelling summary of the stages that the corpus based approach has gone through. In a very illustrative way she distinguishes three phases: ‘the dawn of corpus based translation studies (1993-1995)’, ‘the establishment of corpora in translation studies’ (1996-1999); and ‘the spread of corpora across languages and cultures’ (from 2000 onwards) (ibid.:13). The dawn of CBTS was linked with the creation of big corpora within the corpus linguistics discipline (such as the British

National Corpus – see Chapter 6). The establishment of corpora in translation studies was fruitful in the number of corpora being created as well as in the number of studies taking a corpus-based approach (for instance, the projects included in

Laviosa 1998; other specific projects which directly relate to this research are described in Chapter 6). The third stage in CBTS history saw an expansion of the discipline incorporating approaches from different areas to become more

‘interdisciplinary’ (Laviosa, 2001:23). However the use of corpus methodologies has not become mainstream. The remarkable developments in computer hardware capacity have had a major and positive impact, as it made it possible for single researchers to host large corpora in their personal computers, and the development and availability of tools for the analysis were also beneficial. However, on the other hand, the difficulties in compiling a corpus, which is a time-consuming and expensive task (more in 6.1 and 6.2), to some extent affected the growth of the discipline.

Furthermore, corpus linguistics has also been criticised for focusing too much on the word. It has been argued that language and literature cannot be reduced to computer programmes and statistics: ‘When stylistic features of a text have been transformed into numerical form, they acquire a status that actually prevents them from being perceived as language-for-communication as such’ (Peer, 1989).

However, there have also been calls for a more objective approach to corpus linguistics before it can be considered a science (Ahmad, 2008). It is essential to notice that corpus linguistics and the use of corpora are an ‘important tool in translation studies’ (Aijmer and Alvstad, 2005:1) but these tools need to be complemented with other methodologies and data analysis strategies in order to draw conclusions. As Mona Baker (2004:183) explains ‘in corpus work (…) the real challenge lies in two things: one is how a researcher might select features to focus on, and the other is how he or she might interpret what they find in their data’

(ibid.). This research into the translation of proper names is conducted on this bases, making use of corpus tools in order to facilitate the analysis of larger amounts of data than it could be accomplished manually, but combining quantitative and qualitative methods (for a close critical analysis of the data) with Toury’s methodology for DTS for the interpretation of the results and framing of the findings.

As a response to those issues around corpus compilation, scholars in the field started capitalizing in a much simpler and effective way of collecting texts: the World Wide

Web. ‘The web is immense, free and available by mouse-click. It contains hundreds of billions of words of text and can be used for all manner of language research’

(Kilgarriff and Grefenstette, 2003:1). Ways in which the web can be used as a corpus are currently being explored (Baroni and Bernardini, 2004), (Fletcher, 2004),

(Sharoff, 2006) and there are already tools available which allow to linguistically search the web, WebCorp5 which stores large sections of the web allowing to replicate the searches and presenting the results in the form of concordances is the most popular example (Renouf, 2003). The University of Leeds Centre for

Translation Studies has also been contributing to this field; IntelliText (Wilson et al.,

2010) ‘provides a simple, integrated interface for carrying out a range of searches and statistical analysis on large, publicly available corpora in several languages’

(Kruger et al., 2011:3). In addition, the Centre is also undertaking research projects on Machine Translation (MT) (Sharoff et al., 2009), corpus-based interpreting studies (Peng, 2009) and multimodality (Thomas, 2009). This project on the translation of proper names makes use of these tools developed at the University of

Leeds Centre for Translation Studies, where the corpus of Chick Lit novels is hosted

(details in 6.2.1).

In summary, all these debates together with the growing number of comparable and parallel corpora as well as of CBTS projects (more detail description in Chapter 6) both at collective and individual levels, are contributing to the fast development of the discipline and many scholars join voices with McEnery and Xiao (2008:27) now to predict that ‘the corpus based paradigm will soon enter the mainstream of translation and contrastive studies’ (ibid.).

5 http://www.webcorp.org.uk/

On the other hand, however, there are also calls for a more widespread reach of

CBTS projects, in order to overcome those initial criticisms into the constraint that the single words impose to the discipline. Laviosa concludes her review by calling for more ‘multilingual and interdisciplinary work’ (Ibid., 2011:26). In similar lines

Saldanha’s (2011) recent work with parallel and comparable corpora to uncover the translator’s own stylistics reinforces this point. Saldanha introduces the idea of using metalinguistic information as an answer to some of the patterns shown by her analysis, ‘reflecting the translators’ different conceptualizations of their readership and of their role as intercultural mediators’ (Saldanha, 2011:257). Similarly, a recent volume exploring the current state of the discipline edited by Kruger, Wallmach and

Munday (2011) concludes that ‘all these studies on corpus-based translation studies lead away from a narrow source-text/target-text comparison and towards a broader approach, into areas such as prosody, phraseology, pragmatics, stylistics, discourse and contrastive linguistics’ (Ibid.:7). The future of CBTS lies perhaps in those areas.

1.5 Conclusions

As the present research tries to explain the artefacts of translation through the study of translated texts as compared to their original counterparts, it fits within a Product-

Oriented Descriptive Translation Studies framework, as understood by Holmes

(1988/2000). The analysis is undertaken using tools from corpus linguistics, having electronic versions of the novels allows all the proper names in the texts to be located automatically. In particular, I use named entity recognition programmes for semi-automatic extraction of the proper names and trilingual concordencers for

41 automatic matching of aligned paragraphs in all the languages involved (specific details can be found on 6.3 and 6.4).

Taking a functional approach, each name is analysed in its context, with a view to determining whether the translator chooses to conserve the original name, at the risk of introducing foreign elements into the target culture, or whether s/he opts to substitute the name in an attempt to adapt the text to the new culture, facilitating the task of interpretation for the reader. In other words, this approach follows

Toury’s (1995) initial norm setting translator’s behaviour towards either adequacy or acceptability.

This Chapter has offered an overview of the main Translation Studies theories and concepts used throughout this study. CBTS will be revisited again in Chapter 6, with a detailed description of the corpus tools and methodologies employed in the creation and interrogation of the corpus used in this research.

Chapter 2 moves in a philosophical direction in an attempt to find a definition of proper name which would link the theories of meaning with the theories of the translation of proper names (explored in Chapter 3).

Chapter 2. In Search of a Definition of Proper Name

“The Way that can be told of is not an Unvarying Way: The names that can be named are not unvarying names. It was from the Nameless that Heaven and Earth sprang; the named is but the mother that rears the ten thousand creatures, each after its kinds” Tzu, Lao: Tao Te Ching (p. 1)

2.1 Introduction

Proper names have constituted an important topic of study for scholars interested in studying language since antiquity. The Stoics introduced the difference between common nouns, appellatives for them, and proper names. In his work Cratylus, Plato discusses whether proper names are conventionally assigned or, conversely, naturally connected to the bearer (Sedley, 2003). Those arguments are still nowadays central to the theories of proper names. These theories are divided between those working to understand and describe the linguistic features of proper names (2.2) (see, for instance, Jespersen (1924/1964), Algeo (1973), Allerton

(1987)) and those interested in their philosophical character (Frege (1892), Kripke

(1970/1980), Searle (1958)). These language philosophers interested in proper names have tried to establish whether proper names have semantic meaning or not while working towards setting a theory of meaning in which proper names would play a central role.

The present study is grounded within these discussions around the semantic meaning of proper names. This chapter explores the theories of various language philosophers in search for a theory of meaning (2.3) and uses them to form a

43 definition of proper name (2.4). These theories can be directly related to the problem of the translatability of the proper names, as considered in the literature and tradition of the translation of proper names (Chapter 3). If proper names lack any meaning, the argument goes, then they cannot be translated. But if it is accepted that they have a semantic meaning then scholars talk about the possibility of their translation. Thus, the theory of meaning can be directly linked with the theory for the translation of proper names.

2.2 Linguistic approach to describing proper names.

From a linguistic point of view, scholars work around a definition of proper name based on the description of the main attributes which can be assigned to them and which, in theory, separate them from common nouns. The three criteria traditionally used to define proper names encompass referential, morphosyntactic and semantic features. Algeo (1973) explains them as follows:

Morphosyntactic: Proper names are capitalized Proper names have no plural forms Proper names are used without articles Proper names do not accept restrictive modifiers

Referential: Proper names refer to single unique individuals

Semantic: Proper names do not impute any qualities to the objects designated and are therefore meaningless Proper names have a distinctive form of definition that includes a citation of their expression. (ibid.:9-13)

Otto Jespersen (1924/1964) attempts to describe the actual meaning of proper names from a linguistic point of view, reviewing these traditional criteria. However, he finds shortcomings on their formulations, for instance, he shows how proper names can indeed form plurals in the usual way: ‘in the party there were three

Johns and four Marys’ (ibid.:69) or, with regards to the use of the article, he considers that proper names have ‘an approximation to common nouns’ (ibid.:70).

One example of this approximation is the familiar use of words such as father or mother that are commonly used without the article but to refer to specific individuals. These factors hence lead to the conclusion that the difference between proper names and common nouns is one of ‘degree rather than of kind’ (ibid.:70-1).

Similarly, Gardiner (1954) uses these criteria to form his own definition of proper names which sees proper names as essentially linguistic elements with only one possible referent (ibid.:5). Gardiner sees the different uses of proper names, for example proper names used as adjectives to describe some traits of the bearer (in a sentence such as he is a real Casanova), as ‘perverted’ (ibid.:15) and emphasises that those cases lie ‘outside our immediate problem’ (ibid.). For his part, Sørensen

(1963) brings the Saussurean concepts of signifier and signified to the problem of proper names. He explains that a ‘denotatum’ (ibid.:14) involves the extra-linguistic entities that are referred to by means of a sign. For example, the denotatum of

‘Churchill’ is the flesh and blood entity Churchill. In this way, proper names convey information, and therefore, he concludes, have meaning.

While treatises like Gardiner’s and Sørensen’s are preoccupied mainly with the study of anthroponyms, that is, of personal names, Allerton (1987) introduces a more

45 detailed taxonomy for types of proper names in English and attempts to deal with the syntax of proper names from a systematic point of view. Allerton explores the linguistic criteria traditionally assigned to proper names (as above) in search for the structure that can identify them grammatically (ibid:64). But, he concludes, these grammatical features alone do not form a complete definition of proper name. The semantic characteristics of proper names must be admitted as part of a truly comprehensive definition. As already pointed out in the introduction to this chapter, this particular feature is paramount for the subject of the translation of proper names, as it relates directly with the possibility or impossibility of the translation of proper names (Chapter 3), and as such, the semantic criteria play a prominent role on the definition of proper names used in this study.

2.2 Philosophical approach to the meaning of proper names.

There are two mainstream theories around the semantic meaning of proper names, the referential theories and the descriptive theories. The referential theories were born from Mill’s (1983) ideas which see the proper name as a mere designator of the one specific amongst all of the same type; that is proper names specify one amongst a group of similar entities, for example child > Pablo (a consideration adopted, amongst many others by Gardiner (1954) (as seen above)). This is the most widespread conception of the proper name, latent in definitions in dictionaries and grammars of most languages still today. On the other hand, the descriptive theories take the view that a proper name is associated with certain characteristics of its bearer and these characteristics form the meaning of the name. These ideas developed from Frege’s concepts of proper names having both a sense and a

46 referent and have been followed most importantly by Bertrand Russell (1905) but strongly criticized by Kripke (1970/1980).

The definition of proper name used in this study takes Frege’s concepts as its main basis, but also incorporates referential criteria, as it considers that the attribute of a proper name to point out to one specific entity cannot be disregarded. By taking concepts from both mainstream theories and joining them together into one definition, this study takes an approach to the meaning of proper names that tries to incorporate different aspects of their attributes that are relevant to the translator.

The following sections of this chapter review these theories as well as their main criticisms with a view to understanding the concepts and forging a definition from them.

The first modern theorist to discuss proper names in detail is considered to be John

Stuart Mill (1843), who analysed the differences between common and proper nouns in great detail in his extensive treaty about logic and knowledge, A System of Logic

(1843). Mill understands logic as the study of prepositions, and prepositions are formed by names. Two thirds of the first volume of this theoretical work are devoted to names, mostly common names but also to proper names. Mill introduces the idea that proper names only denote, they cannot connote. Through the use of a proper name we are not offering any information or characteristics of the bearer other than its name. He uses the name of the city York as an example (ibid.:44): when using the name York, the speaker is not explicitly saying that the city has a minster. The listener may know, however, that it does, but that is because of other previous

47 knowledge that the listener may have, not because it is integral to the name. In this sense, Mill considers proper names as ‘unmeaning marks’ (ibid.:43).

For his views of proper names as mere referents, Mill (1843) soon became known as the father of the notion of proper names lacking of any meaning and his theory is famously distilled down to his words describing proper names as ‘empty labels’

(ibid.:43). These simplified versions of his ideas were easily rebutted, as the following example, used by Bernárdez (1987:16), highlights:

si hay un personaje llamado Albert que vive en la Fifth Avenue de New York podríamos decir que Alberto vive en la Quinta Avenida de Nueva York pero también, en principio, podríamos hacer una sustitucion completa y decir que Torcuato vive en la Calle Mayor de Fregenal de la Sierra. Si los nombres propios son meras etiquetas, ¿por qué no? (ibid.)

This makes Mill’s theory appear invalid. However, when Mill is studied in more detail, it becomes clear that this criticism only takes part of his theory into account, and that Mill also voices some very valid points. As is shown by his example of the name of the city York (1843:44), simply by uttering this name the speaker is in no way implying certain characteristics of this city; however, due to the previous knowledge that the listener may have, he may immediately connect name and characteristics in his mind. But this is due to previous knowledge, not to anything implied by the proper name of the city.

On the other hand, in 1892 Frege’s theory of meaning introduced a different view which opposed Mill’s idea of proper names as mere referents. Frege (1892) sees proper names as having both a referent and a sense. The referent is the actual thing that is being designated by the name and the sense is the cognitive content associated with it. Frege uses the example of the planet Venus (ibid.:24), which sometimes is referred to as the Morning Star and sometimes as the Evening Star.

These two expressions have the same referent but different senses. For Frege ‘a proper name (…) expresses its sense, stands for or designates its referent’

(ibid.:27). Competence in a language is what makes the users able to grasp the senses of the names. Frege stresses the difference between the sense of a name and the idea that the use of such name arises in the mind of the listener, a point that can be linked to Sørensen’s (1963) ideas of denotatum versus meaning, as posited above.

Directly related to these concepts are Russell’s (1905) ideas which see proper names as descriptions; through those descriptions we are able to form ideas even of things we are not acquainted with. Russell sets out his theory of meaning in the article ‘On

Denoting’ (1905), where he introduces the concept of denoting phrases. A denoting phrase is formed by a determiner (such as ‘the’, ‘a’, ‘my’, etc.) plus a noun phrase, for example the king of Spain. A denoting phrase can be either a definite description, if the determiner is the definite article (or a possessive adjective) or an indefinite description if the determiner is the indefinite article. In his work of 1912, The

Problems of Philosophy, Russell considers proper names as definite descriptions ‘in disguise’ (ibid.:54-59). When something is said about Julius Caesar, Russell explains,

49 the speaker and listener cannot have in mind Caesar himself; instead what is in their mind is some description of him which is familiar to them, for instance famous

Roman Emperor. According to Russell, the sense of a proper name is set by the descriptive information that is believed to be true of its referent.

Searle (1958, 1967) takes Russell’s ideas further proposing that proper names are tied to a series of descriptions that can always be applied to them and which serve to define each name: ‘Though proper names do not normally assert or specify any characteristics, their referring uses nonetheless presuppose that the object to which they purport to refer has certain characteristics’ (1958:170-171). According to Searle

(1967:490), the referent of the proper name is not determined by one description but rather by a group of them. ‘For example, Aristotle was a Greek; a philosopher, the tutor of Alexander the Great; the author of the Nicomachean Ethics, the

Metaphysisc, and the De Interpretatione; and the leader of the school of the Lyceum at Athens’ (ibid.). Hence, Searle reinforces the idea that proper names have both a referent and a sense, while giving strength to the sense which forms the meaning of the name. This meaning, of course, changes with the different associations gathered by the entity, for instance, Julius Caesar, famous roman Emperor or demised of the

Roman Republic. The importance of Searle’s theories lies in stressing the coexistence of both a sense and a referent, without such a clear cut divide between denoting and connoting; both are present in a proper name. This is key for this study (see below, and Chapter 7) and as such this idea is best illustrated with an example from the corpus: at a hen party all the ladies are talking about their aspiration to be like

Kate Moss; this one name is both denoting their favourite model as well as

50 connoting some of her main attributes (extremely beautiful, slim and rich). Whence,

Searle reinforces the idea that a name acquires the characteristics of its bearer during its life and these characteristics form the meaning of the name.

Kripke is the main detractor of these theories which he criticized in a series of lectures at Princeton University in 1970. Kripke considers proper names as ‘rigid designators’ (1970/1980:49) in all possible worlds. He argues that a person called

Aristotle would still be called Aristotle even if he had not done all the things that have been attributed to him, and in this sense, proper names are rigid. If a person goes into a room and meets Mr Scott for the first time, s/he can still talk to and acknowledge Mr Scott by his name, even if s/he does not know any of these descriptions that are attributed to him. In this sense, Kripke is closer to Mill’s (1843) view of proper names as mere referents. In fact, for Kripke, proper names are ‘rigid designators in all possible worlds’ (ibid.:49). Kripke uses the example Nixon, president of the US, to explain that Nixon would have still been called Nixon even if he had not become president, and in this sense, he sees names as rigid, independent of the characteristics of their bearer.

So, while Russell (1905) differentiates between proper names, Elizabeth II, to definite descriptions, the Queen of England; and considers that the meaning of a name is composed of the different definite descriptions that it generates, Kripke

(1970/1980) clearly disagrees with this, as he considers that proper names do not have any meaning at all, as they exist even without these associations, or rather besides them.

All these theories taken individually seem to pose accurate descriptions of the behaviour and semantic attributes of proper names; however, they do not seem to stand together when they are grouped. The missing argument is presented by

Jespersen (1924). Jespersen accepts that it is impossible to tell characteristics from a name such as John; in that sense proper names act much as Mill views them, as

‘unmeaning masks’ (Mill:1843:43). However, from a semantic point of view,

Jespersen maintains that ‘proper names connote qualities by which the bearers of the names are known’ (1924/1964:71). Crucially, Jespersen considers that Mill failed to see the contextual and situational value of proper names as they are actually used by speakers of a language. When John is used in a specific situation, the speakers understand the referent and understand the meaning and characteristics of the bearer. In that sense, proper names have even more semantic meaning than common names, as they ‘connote the greatest number of attributes’ (Jespersen,

1924/1964:66). Take the common noun bear, defined as a ‘mammal with thick fur and short tail’ (Oxford dictionary). But more specifically there is, for instance, the very famous Paddington Bear, that little teddy with red hat and wellington boots that we are all familiar with. When we say Paddington Bear we all think of the same one bear, and in this way the name refers to its one referent. But what happens, for example, in a family where the little toddler has named her favourite toy teddy after her favourite cartoon character, Paddington Bear? For that particular family, the name Paddington Bear would take them straight to their daughter’s teddy in the first instance and not to the cartoon character. This highlights the importance of the context and the shared knowledge of the speakers when it comes to assigning meaning to a given name. These ideas can be linked to Relevance Theory in

Translation (Gutt, 1998, Gutt, 2000). Gutt uses concepts from Relevance Theory

(Sperber and Willson, 1995) to explain that the meaning of an utterance ‘depends, not only on its semantic content, but crucially on the context in which it is interpreted or, more technically, on the contextual information with which it is inferentially combined’ (Gutt, 1998:42). In this sense, the family of the toddler who owns a teddy bear called Paddington all share a context where this name has one specific referent. Applied to a translation setting, the translator has to interpret this context for the target audience, when this context is not shared by the SL and TL audiences.

In these lines, Jespersen introduces the context and the situation (that is, when a name is actually being used in a linguistic exchange) as parameters that should be taken into account when considering proper names. Those arguments place

Jespersen between the Millian-Kripke view of proper names as conveyors of no information about the bearer and the Fregean-Russell view of proper names as descriptors, with a sense as well as a referent. Thus, in isolation, the name York says nothing of the city it is naming, but in the context in which it is used it can be presupposed that it means something for both speaker/writer and listener/reader.

True as well that Nixon would have been called Nixon even if he had not become president, but in the history of this world he did indeed become president and when

Nixon is being used to refer to this one man in particular, it cannot be denied that he was president once, and this fact goes hand in hand with the name Nixon. So, it can be said that Frege’s ideas pointing out that names have both a sense and a referent were accurate, and I would go even further to argue that this referent is fixed within

53 the context and situation in which the name is being used; a name is used at one particular time to refer to one particular object or individual, an in that situation and context the characteristics, or ‘definite descriptions’ to use Russell’s (1905) words, which accompany the referent behind the name are all relevant. So, all the statements are correct in their isolated theories, however, proper names rarely occur in isolation, and it is this context and situation which are key, for the theory of meaning as much as for the theory of the translation of proper names as it highlights the difficulties that lie behind the recreation of this context for a different target audience (Gutt, 1998).

From this standpoint, it should now be possible to form a definition of proper name which takes into account the attributes of proper names more relevant for a translation setting. According to Mill (1843) a proper name is a word that identifies its object. According to Frege (1892) this word as well as a referent, its object, has a sense, what this object represents. Meanwhile, Jespersen (1924/1964) incorporates the idea that it is the context and the situation in which a proper name occurs where this synergy between the referent and the sense takes place and acquires meaning.

In addition, Frege (1892), Gardiner (1954) and Mill (1843) also indicated that a name can be used to denote or connote. On this point, however, the widespread view seems to be not to consider connotative names as proper names, and as such they are therefore left out of most treatises on proper names (for instance Gardiner

(1954) or Zabeeh (1968)). But this study lies closer to Searle’s (1967) notion discarding a rigid referent-sense/denotation-connotation approach. All names have a bit of both in them, some are more connotative than others, but connotations are an

54 intrinsic part of proper names, as is referring to one particular entity. This point has already been addressed earlier in this chapter and will again be highlighted in

Chapter 7 of this study, and most specifically in 7.3.2 with regards to a classification of proper names from a semantic point of view. The categories envisaged on this classification attempt to encompass this dual aspect of the names (in between denoting and connoting) always at work within a proper name in the context in which it is being used. The examples of names in context used throughout this study and in particular those highlighted in section 9.4 of the Results Chapter, also illustrate this idea.

2.4 A definition of Proper Name

On account of the arguments explored above, I will define a proper as follows: a word, or group of words, that identifies an individual entity setting it apart from all the other entities in the same class and which, in the context and situation in which it is being used, has a specific referent and sense. Proper names both denote this referent and connote the characteristics which form its sense.

Proper name: a word, or group of words (Gardiner), that identifies (Mill) an individual entity, (Cantora) setting it apart from all the other entities in the same class and which, in the context and situation in which it is being used (Jespersen), has a specific (Cantora) referent and sense (Frege). Proper names both

(Cantora/Searle) denote this referent and connote (Frege/Gardiner) the characteristics which form its sense (Russell).

The core component of this definition is Frege’s concepts of sense and reference together with Searle’s notion disregarding a rigid denoting/connoting approach.

Jespersen’s key idea of the context and situation as paramount is also crucial in this definition, as is the idea that a proper name stands first and for all to name one individual entity setting it apart from all of the rest in its class. In summary, the key points in this definition then, see proper names as referential entities with specific connotations associated to this one referent.

2.5 Conclusions

This Chapter has reviewed the theories of meaning with a view to composing a definition of proper name. Both linguistic and semantic criteria and theories are considered, and the different ideas posed throughout the years are taken into account and amalgamated in one united definition of the object of study of this project. The theories of meaning explored in this Chapter relate directly to the arguments generally used in the discussions around the translation of proper names.

Chapter 3 now considers the relationship of proper names to translation theories and reviews the most salient works on the subject.

Chapter 3. Proper Names and Translation

“You’ll see that Argentine women carry themselves like no other, they seem to be more worldly, perhaps because they are all psychoanalyzed, or perhaps because of those names and last names they are given. You have to admit that it’s one thing to be called Mariluz Padilla Soto, and quite another to be named Carlota, Carlota Fainberg” Muñoz Molina, A.: Carlota Fainberg

3.1 Introduction

That ‘“proper names are never translated”’, Christiane Nord (2003:182) observes, is a ‘rule deeply rooted in many people’s minds.’ However ‘looking at translated texts we find that translators do all sorts of things with proper names’ (ibid.). This clash between perception and reality emphasises the importance of studying proper names in a translation context, as the view generally associated with proper names

(e.g. that they do not change in translation) does not consistently reflect the reality of translators’ work with proper names. This chapter is devoted to exploring different studies and theories around the translation of proper names to position the present project amongst the literature on the subject and to highlight the different approaches taken both to translate proper names as well as to study translated proper names.

The literature on the translation of proper names has followed a similar pattern to that of translation studies in general, starting traditionally from prescriptive approaches observing detailed norms for the different types of names and what the translator should do with each category (Levy, 1965, Santoyo, 1988, Newmark,

1981), and progressively moving onto more descriptive methodologies, tackling the

57 problem from observation of the treatment given to proper names in real translated texts (Franco Aixelá, 2000, Moya, 2000, Cuéllar Lázaro, 2000, Viezzi, 2004).

Traditionally, the argument was centred around the possibility or impossibility of the translation of proper names. As stressed in Chapter 2, these ideas can be directly linked to the semantic theories of proper names posited by language philosophers.

The translatability or untranslatability of proper names was discussed in the light of their meaning or lack of meaning. If proper names are considered to be meaningless entities, there is no need to engage in their translation; on the other hand, if proper names do have meaning, they should be translated.

3.2 Arguments against the translation of proper names

Amongst those who do not believe in the translation of proper names we find the theories of Zabeeh, Newmark, Manczak and Muñoz Martín. These will be briefly summarized below in order to show the various arguments for untranslatability.

In a comprehensive volume devoted to the semantics and pragmatics of proper names, Zabeeh (1968) reviews the different semantic and linguistic theories related to the subject, touching as well upon the problem of their translation (ibid.:69-71).

Zabeeh sees proper names as ‘international items which belong to no specific language’ (ibid.) and believes that they cannot and should not be translated. He admits, however, that different positions should be considered for proper names which originally were ‘meant to be descriptions’ (ibid.), for instance in a sentence

58 like ‘he is like Superman’. He states that the translator should translate those expressions which are ‘intended to say something about their bearers’ (ibid.).

In his first translator training course, Approaches to Translation, Newmark (1981:70) declares proper nouns to be ‘outside language’ and to belong to the ‘encyclopaedia not to the dictionary’, with ‘no meaning or connotations therefore, both untranslatable and not to be translated’. However, later on, (1988) in A Textbook on

Translation, he includes a whole chapter on the subject, analysing different translation techniques and possibilities for each different category of name (due to its relevance for this study Newmark’s categorization of translation techniques is studied in detail in section 4.3.1).

Manczak (1991:26) discusses the difference between proper names and common names on the grounds of the impossibility of the translation of proper names. He defends this impossibility of the translation based on statistical evidence obtained in his study on the translation to Polish of proper names excerpted from the French newspaper Figaro. Manczak establishes that only 4% of the proper names studied were subjected to any form of translation, therefore proving, in his opinion, the statistical unlikelihood of the translation of proper names.

In his review of the major current linguistic theories related to translation, Muñoz

Martín (1995:194) touches upon the subject of translating proper names. He cannot admit the possibility of the translation of proper names as they are terms with no meaning. He only accepts the repetition or phonetic or orthographic transcriptions in

59 order to adapt the names to the alphabet of the target language as admissible techniques for their translation.

The problem of transferring the names to and from languages with different alphabets has been studied in more detail by Teresa Espinal (1989-1991:74-5). She considers that for languages where the alphabet is different, the translator should use transliteration, that is, the name should be re-written in the new language following the alphabet of that language. This should be applied for example to languages like Chinese, Russian or Arabic. This method to transfer proper names is also known as naturalization (Newmark, 1988:82). The term used for this procedure in the present study is transcription (8.2.4) which Espinal explores in relation to fictional names which have been invented by the author. For those names, she explains that transcription should be used, giving special attention to the original pronunciation in order to try and replicate it in the language of the translation. She uses as an example the Catalan translations of Tolkien’s names in The Lord of the

Rings:

Gandalf – Gàndalf

Gimli – Guimli

Gamgee – Gamgí

These scholars’ attempts to prove the impossibility of the translation of proper names all give concessions in certain situations; thus, for them translation should be considered, either when different alphabets are involved or with connotative names.

These points undermine the absolute untranslatability of proper names and thus move the subject towards consideration of different possibilities for the translation of proper names.

3.3 Arguments in favour of the translation of proper names

The theories of the scholars that follow the possibility of the translation of proper names (for instance, Levy, 1965, Bernárdez, 1988, Moya, 2000, Nord, 2003, Viezzi,

2004) were originally prescriptive, and focus on establishing set rules for their translation; for example, Levy accounts for three different possibilities:

For proper names the translator can use: Transliteration: Mr Ford – Mr Ford Substitution: Mr Newman – Herr Neumann Loan Translation: Everyman – Jedermann (Levy, 1965:81)

The scholars interested in studying the translation of proper names cover a varied number of schools of thoughts and countries, for instance in Germany, Kalverkämper

(Kalverkamper, 1995) links the subject of proper names with Lexicography at the same time that he reviews the latest research on the subject while Lietz (1992) concentrates in particular on the translation of literary proper names.

As the research reported here focuses on the translation of British proper names into Spanish and Italian, in this section I will, therefore, concentrate on studies originating from those countries.

Proper names have always been a very strong subject in the Spanish tradition within the TS discipline. Several centuries ago, it was common practice in Spain to translate all the names and it was not unusual to find in Spanish texts a Juan Sebastián Bach and even a Guillermo Shakespeare (Cuéllar Lázaro, 2004). In keeping with this tradition scholars were originally preoccupied with setting prescriptive rules to guide translator behaviour around proper names. In 1983 the central discussion panel of a major translation conference in Madrid6 was occupied by these issues. Prominent

Spanish scholars joined together to discuss ideas around the subject. For example,

Santoyo (1988) takes the idea that the linguistic tradition of each country has already decided which names are translated and which are transliterated or copied and lists different types of proper names and what tradition rules for each: for instance, names of buildings, restaurants or newspapers are copied, whereas book titles, continents, countries or geographic features are translated. The main point that Santoyo stresses is the need for the translator to be aware of the established form for the SL name in the TL and to respect this form. Santoyo only considers problematic the literary names chosen due to specific features of the name by the authors. Similarly, Bernárdez (1998) considers the idea of connotative names and the different possibilities for the translator. According to him, each name should be considered singularly, taking into account the function in the ST and aiming to avoid introducing connotations to the TL which were not present in the original, but at the same time, the connotations latent in ST should not be cancelled in the translation.

Only considering each name individually and independently can this be achieved. A similar point is presented by Hernández (1998), who also introduces the idea of

6 Conference Problemas de traducción, organised by the Fundación Alfonso X el Sabio in Madrid November 1983 62 adapting the name to the target readers always without losing the exoticism of the original name and taking into account the established form for the name in the target culture. Similarly, Cantera (1988) considers French toponyms and anthroponyms and how tradition has treated them in their transfer into the Spanish language. All in all, these scholars agree on two main points: firstly, the importance of the established linguistic form of the SL name in the TL and, secondly, the special considerations that need to be given to connotative names.

These prescriptive considerations gave way to more descriptive approaches from the mid-1990s onwards, with the growing popularity of DTS (Toury, 1995. See Chapter

1). A number of scholars took to comparing translated texts in search for the real treatment of proper names in real translations. One of the most prestigious and cited studies on the subject is the one conducted by Virgilio Moya in 2000. Moya studied over a thousand foreign proper names extracted from Spanish newspapers published during the 1990s. He presents a descriptive analysis with the aim of mapping the actual treatment that foreign proper names receive in the Spanish media. Moya accepts repetition or adaptation as valid techniques for the translation of proper names, he never considers the impossibility of the translation of proper names. He concludes that the current tendency points towards borrowing the names in their original forms.

Interestingly, contrary to the views expressed by the traditional scholars during the

1980s, in his study Moya observes a certain tendency to recover the original given names of famous figures or characters which had previously been adapted to the

63 language. Charles Dickens and Martin Luther are examples of foreign names that were translated in the past (Carlos Dickens and Martín Lutero) but are now once again used in their original forms. However, other names, such as those of certain royal figures or that of Pope John Paul II are found to be kept in their traditional

Spanish form (el Príncipe Guillermo or Juan Pablo II). Although Moya’s descriptive analysis shows that this is what is actually happening in translation, it is a controversial point as the prescriptive theories of translation of proper names had always claimed the necessity to respect tradition, in the words of Santoyo ‘it is not up to the translator to change the names that are established or to choose new ones, as it is the tradition that rules’ (1988:45) [my translation].

An example of what happens when the translator does not follow the convention is presented by Lefevere and Bassnett (Lefevere and Bassnett, 1990) when they recount the story of Proust’s grandmother’s reaction to the changes of the names in the version of the Odyssey as she traditionally knew it:

Proust’s grandmother likes the translations she has grown up with. ‘The’ Odyssey for her is a translation in which the hero is still called by his Latinised name: Ulysses, and in which the goddess Athenea is likewise still called Minerva. Other Odysseys or rather, other texts deemed to represent Homer’s Odyssey, simply will not do, they are impostors. (ibid.:1990:2)

Other translation scholars, however, join voices with Moya. For example Anthony

Pym (1992:205) states that the translator should ‘not feel obliged to respect tradition for the sake of tradition. Why should the Spanish ”Carlos Marx” not become

”Karl”?’. Like Moya, Pym does not accept the untranslatability of the proper name:

‘an apparently untranslated term’, Pym explains, ‘in an otherwise translated text must be received as a term which has at least been processed translationally’

(ibid.:73). For Pym, when the translator is taking the decision to retain or repeat a proper name, s/he is undertaking proper ‘translational work’ (ibid.:75). A similar point will be posited later on in this study (in 4.3.5 as well as in 8.3) where it will be argued that translator intervention occurs at all levels of the translation process. The idea of the untranslatability of proper names stems from the view that translation techniques such as repetition or transliteration are not considered acts of translation.

This argument has featured prominently in the theories of the scholars who do not believe in the possibility of the translation of proper names seen in section 3.2

(Zabeeh, 1968, Manczack, 1991 or Newmark’s, 1981 ideas on the subject). In a similar way, as seen before, this view has also been sustained by the classical

Spanish theorists, for instance Vázquez (1988:39) emphasizes that they cannot be called translations, but merely phonetical adaptations.

Franco Aixelà (1996, 2000) is another modern Spanish theorist who strongly believes in the possibility of the translation of proper names, and as such he conducted a study of real translated proper names. Franco Aixelà (2000) analysed over 11,000 proper names translated into Spanish from English sources. The genres explored in this study include modern classics, children literature, thrillers, romances, poetry, essays and journalism and range through a period of 200 years, with Blake’s Songs of Innocence being the eldest text (1789) and Wildcat by

Rebecca Brandewyne (1995) the most recent one.

Franco Aixelà captured the different strategies employed by the different translators with an aim to describing how translators work as well as the reasons that motivate their approaches. His model of translation strategies is divided into conservation or substitution, depending on the degree of conservation of the name in the SL or its adaptation to the new TL readers and culture. He identifies a total of 12 different strategies, 6 within each block (for a detail description of these see 4.3.3). Franco

Aixelà’s model features heavily in the map of procedures devised for this study and as such is explored in great detail in 4.3.3 as well as all throughout Chapter 8.

In his conclusions, Franco Aixelà finds that proper names can and indeed are translated. He considers, like Pym and Moya, that repetition (when the proper names is left in their original form) is an act of translation on its own account, so the fact that it is the most common strategy does not affect this conclusion. In addition, he also explores other factors that condition the translator when deciding on the most appropriate translation. Those factors include the purpose of the translation, the target reader’s expectations as well as the genre or, for literary texts, the commercial aims of the publishing company.

On the lines of descriptive projects like Moya’s and Franco Aixelá’s, the most relevant study of proper names taken from the Italian tradition has been undertaken by

Viezzi (2004). His work with proper names represents one of the more salient and cited approaches to the subject in his country. Viezzi conducted a study of real names as well as fictional names and titles of works of fiction and films from a

‘translinguistic and transcultural point of view’ (ibid.:11 [my translation]). In his

66 account of translated titles Viezzi stresses how semantic correspondence is often ignored in favour of other factors, such as producing a new title which will be viable commercially in the TL culture (also similar to Franco Aixelá’s conclusion seen above). With regards to personal names, animal names and places, he points out that they are not mere labels and that they often incur the use of specific procedures for their translation. The strongest point of this work is to bring proper names out as specific objects of study with enough relevance and content in his country and succeeded in inspiring similar studies across other text types. For instance, Loiacono (2011) uses his definition and account of attributes of proper names to place legal terminology at the centre of a study on culture-specific elements and he successfully applies the procedures identified by Viezzi to a set of legal documents concluding that legal terminology often acts in a similar way to how proper names do.

At the same time as these major investigations were taking shape, a number of similar attempts at categorising and describing the techniques employed to translate proper names started to emerge on a smaller scale. These include reviews of literature related to the question (Mizani, 2008, Cuéllar Lázaro, 2004) or the procedures employed to translate one particular book (Dastjerdi, 2008) (García

García, 2001), or across the novels of a particular author (Gutiérrez Rodríguez,

2004). Children’s literature is also a relevant genre which has generated considerable interest (Nord, 2003, Manini, 1996, Jaleniauskienė and Čičelytė, 2009), as it is generally considered, and empirically shown, that proper names are rarely borrowed when the target recipients are children; the classic fairy tales are perhaps

67 the best known and more illustrative examples of these, Cinderella is known as

Cenicienta in Spanish and as Cenerentola in Italian, Snow White as Blanca Nieves in

Spanish and Biancaneve in Italian or The Little Mermaid as La Sirenita in Spanish and La Sirenetta in Italian, for instance. However, this might also be starting to change, as similarly to what has been happening with classic names (such as the

Charles Dickens case described above), there seems to be more variety in the translation procedures employed with names in more modern children literature.

Harry Potter is the most representative example of this new tendency. The vast range of proper names in these novels, used in many cases to illustrate some special trait of the character (Slythering like a snake), are sometimes adapted to the new language (Slythering is Serpeverde in the Italian versions, serpe meaning nasty or mean) but other times borrowed across in their original form (as is the case in the

Spanish versions where little changes take place emphasising therefore their foreignness). This points as well towards important differences between the tendencies shown by translators from different countries; as the previous examples suggest, the Italian versions of these novels seem to show TL orientation. The names in the translations of the Harry Potter series is a subject that features heavily in the latest literature on translated proper names (Penrod, 2010, Willems and

Mussche, 2010). The work of Fernandes (2004), is perhaps the most exhaustive of these studies. Fernandes takes a corpus-based approach to the study of the translation into Brazilian Portuguese of the proper names in the Harry Potter series as well as in the Chronicles of Narnia. Fernandes uses a parallel corpus containing the SL texts and their Brazilian translations to explore the different ways in which the names are translated. He concludes that translators do many things with proper

68 names and that the use of a parallel corpus might reveal practices that would never been thought before (cf. Fernandes, 2006).

The value of these studies lies in consistently highlighting proper names as a relevant subject of study. They also add strength to the discipline with new approaches and new interpretations and evidences to complement (and confront) the classic idea that ‘proper names are never translated’ (Nord, 2003) which opened this chapter.

There is, however, one issue common to all the scholars in favour of the translation of proper names as much as to those against it; they all believe in the special treatment that should be accorded to a proper name which is performing a special function in the ST. Or in the words of Zabeeh (1968:69), those proper names which are not being used ‘purely as identification marks’, should be considered carefully.

For these names some authors consider that the translator should aim to keep the same function and connotations for the proper name in the TL culture. For example, in an article devoted to the translation of literary proper names, Martinet sustains that:

Pour le traducteur, il s’agira d’abord de savoir si un nom propre donné est donotatum, indice de culture ou les deux à la fois. De savoir ensuite s’il es porteur d’un faisceau de connotations spécifiques ou d’un réseau d’associations propres à l’utilisateur. Et enfin, dans la langue cible, de choisir des éléments linguistiques qui sont aptes à susciter des associations analogues à celles que peut avoir eues le lecteur de l’ouvrage original. (Martinet, 1982:398)

Similarly, Newmark (1993:70) considers that ‘the attempt must be to reproduce the connotations of the original in the TL, but to find a name consonant with SL nomenclature, thus preserving the character’s nationality’.

For his part, Cuéllar Lázaro presents a possible method of translating literary proper names:

En el momento en que un traductor se enfrenta a un nombre propio, debe considerar en primer lugar si éste tiene o no contenido semántico, debe después considerar si dicho contenido puede ser vertido a la lengua meta y una vez traducido debe considerar si evoca el mismo referente en la nueva lengua que en la lengua origen. (Cuéllar Lázaro, 2000:114)

Cuéllar Lázaro highlights, like Martinet, the importance of carefully considering each name and its semantic content, if any, and how best to transfer this across to the TL.

Interestingly, Cuéllar Lázaro suggests as well looking back at the translated term to assess to what extent that semantic content is replicated in the TL version.

This method suggested by Cuéllar Lázaro is, to a certain extent, applied in the analysis of names in context conducted in 9.4 of this study, where a selection of names from our corpus will be placed in the context in which they occur in the ST original and the translation will be considered from a cultural and semantic point of view, in a similar way to that suggested by Cuéllar Lázaro. Likewise, a procedural approach to the analysis of the techniques employed in the translation of all the

70 names extracted from the corpus, not just for connotative names, will be undertaken following the path settled by Newmark, Moya and Franco Aixelá.

3.4 Conclusions

This Chapter directly links the theories of meaning to the theories of the translation of proper names. Scholars interested in the translation of proper names have argued the translatability or untranslatability of proper names on the grounds of their lack or not of meaning. Chapter 2 revised the different theories of meaning around proper names, and concluded with a definition where proper names acquire a set referent and a set sense when used in a set context. From this consideration, proper names clearly become translatable entities; an approach that differs from the theories commented on in section 3.2 of this chapter, which argued for the untranslatability of proper names mainly on the grounds of their attributes as denotative elements. Taking proper names as connotative elements (as in Chapter

2) clearly positions this study within the theories which acknowledge the translatability of proper names. If translatability is accepted, the question then turns to the procedures that are used to achieve it.

The present thesis will explore this question along the lines of projects such as

Franco Aixelá’s and Viezzi’s above. The novelty of this research, however, lies precisely in considering proper names as connotative elements in a set context, and the value added comes from incorporating this context and the specific role that the proper name is playing, to the linguistic research.

The theoretical and practical relationship of proper names to translation procedures is essential for this study. Chapter 4 is devoted to it, examining different studies of proper names and translation procedures. The aim will be to introduce the key elements which inspire and inform the model of translation procedures (Chapter 8) that will be used to analyse the proper names in the corpus of Chick Lit novels created for this study.

Chapter 4. Translation Procedures and Proper Names

“I locked myself in a hotel room in the World Trade Centre, and I wrote for a week, adding material for Americans who might not know where Oxford St. was or what you’d find if you walked down it” Gaiman, N.: Neverwhere. Introduction to the Author’s preferred text edition.

4.1 Introduction

The analysis of the procedures used by translators to transfer a text from SL to TL has been a paramount topic in Translation Studies since the 1950s, when Vinay and

Darbelnet presented their methodology for translation based around a comparative stylistics of French and English (1958/1995). This methodology constituted the first concerted attempt at classifying and describing the different options available for translators. Since then, the subject has been much discussed, initially from a more linguistic and prescriptive point of view and progressively moving, with the shifts in interest in Translation Studies, to more descriptive approaches. This culminated in ‘a rise in interest in empirical research into translation procedures in the 1990s’

(Kearns, 1997:283).

Vinay and Darbelnet’s (1958/1995) ‘classical taxonomy of linguistic changes in translation’ had ‘a very wide impact’ (Munday, 2012:85). Their model comprises two overall strategies and seven procedures within them. The distinction between strategies and procedures is crucial. Munday disambiguates the terms explaining that ‘a strategy is the overall orientation of a translated text while a procedure is a specific technique used at a given point in a text’ (ibid:22).

This chapter introduces the theoretical approach to the mapping of translation procedures proposed in this study for describing the techniques used by translators with proper names (described in detail in Chapter 8). This map is drawn from previous models, taking Vinay and Darbelnet’s taxonomy as the starting point but considering more closely aspects from other approaches designed specifically for the translation of culture-specific elements. Hermans (1988) starts by drawing attention to the role that proper names play in literary texts and how they can be used to infer translation norms; Franco Aixelá (2000) considers the way in which translated proper names adapt to the new target culture and Leppihaulme (1997) adds a different dimension proposing a cultural analysis of the names both in ST and TT.

Finally, Marco’s (2004) idea of placing the procedures on a continuum opens up the possibility of measuring the level of changes to which the names are subjected. All these taxonomies and studies, together with Newmark’s (1988) seminal description of procedures for translating culture, are taken into account and incorporated to the mapping of translation procedures used to classify the techniques employed to translate the proper names from our Bridget corpus of Chick Lit novels.

The present chapter starts with a review of Vinay and Darbelnet’s original taxonomy as well as an overview of similar approaches both from Spain and Italy. It then moves to describe in detail the approaches combined and integrated to create our map: Newmark (1988), Hermans (1988), Leppihalme (1997), Franco Aixelá (2000) and Marco (2004).

4.2 Vinay and Darbelnet and the rise in empirical research into translation procedures

In Vinay and Darbelnet’s taxonomy, the strategies relate to ‘direct’ or ‘literal’ translation (see A below) and to ‘indirect’ or ‘oblique’ translation (see B below), when literal translation is not possible. The nature of the process of translation is fundamentally different within each strategy – direct translation encompasses the procedures of borrowing (A1), calque (A2) and literal translation (A3), whereas oblique translation covers transposition (B1), modulation (B2), equivalence (B3) and adaptation (B4).

Strategy A) Direct Translation Procedures:

A1) Borrowing – use of the same foreign word in the target language. The use of this procedure can sometimes lead to the introduction of new elements into the target culture.

A2) Calque – ‘a special form of borrowing’ (Vinay and Darbelent, 1958/1995:85) where the foreign element is borrowed but each component is translated literally.

A3) Literal Translation – word - for -word translation. The translated term segment must conform to the grammatical conventions of the TL.

Strategy B) Oblique Translation Procedures:

B1) Transposition: change in the grammatical category but without undertaking a change of sense

B2) Modulation: change in the point of view, which implies a change in the message.

This procedure is used when literal translation would be correct grammatically but the resulting segment would not be natural in the TL.

B3) Equivalence: changes, usually stylistic or structural, which affect the message making it better adapted to the target culture. This is used mainly with idioms or proverbs.

B4) Adaptation: When the reality referred to in the ST does not exist in the TL culture, an adaptation is used in order to create an equivalent situation with more validity in the TL culture.

Following Vinay and Darbelnet, other authors have presented similar taxonomies to describe the process of translation. In Spain, for instance, the most exhaustive and influential classification of translation methods and processes is presented by

Vázquez-Ayora (1977). Along similar lines, García Yebra’s (1982) description of the most common translation procedures available to translators included in his manual of translation is also commonly cited. Their main aim was to describe the specific changes which take place when translating from SL to TL and were successful at structuring and describing the field of Translation Studies. The aim of these models

76 was to offer a terminology to describe the translation procedures generally employed by translators to get from ST to TT, whereas from the 1990s onwards there was a shift towards the study of procedures employed by translators with specific translation problems (as opposed to describing the possible methods hypothetically available to them to translate the text at all levels). On these lines, Section 4.3 below reviews some approaches taken to the study of culture-specific elements.

4.3 Translation procedures for culture-bound elements

The development of DTS and the growing popularity of Toury’s methodology (Toury,

1995) during the 1980s and 1990s (see Chapter 1), saw an expansion in the number of comparative studies, which applied Toury’s methodology to analyse specific pairs of SL vs TL. This brought about a specialization of these new empirical studies, which shifted the attention towards the description and categorization of the procedures employed with specific translation problems, such as the translation of metaphors, wordplay, or of language for specific purposes, for instance (Florin,

1993), (Molina, 2006), (Delabastita, 1996), (Mayoral, 1999/2000).

As has been indicated in previous Chapters (Chapter 2, 3), proper names are connotative as well as referential items, and this makes them a prime subject for the investigation of translation procedures. Amongst all of the studies in the area (see

Section 3.3 above) the most relevant studies for this project include Newmark’s

(1998) categorization of procedures for translating culture, Hermans’ (1998) work linking proper names and translation norms, Leppihalme’s (1997) approach to names

77 in context and Franco Aixelá’s (2000) and Marco’s (2004) detailed accounts of translation procedures used with culture-specific elements in real translated texts.

Each one of these proposals is examined below, as they form the basis of the model for translation procedures used with the names in BRIDGET.

4.3.1 Newmark (1988)

Newmark originally envisaged proper names as untranslatable items, he also proposed a method for transferring them from one language to another (as seen in sections 3.2 and 3.3). This apparent contradiction comes to show, perhaps, the extent of his comprehension of proper names as a translation problem. In Newmark

(1988) he devotes a whole chapter to the subject of translation and culture, where he also includes special notes on the translation of proper names. He defines culture

‘as the way of life and its manifestations that are peculiar to a community that uses a particular language as its means of expression’ (ibid.:94). Under this definition, proper names immediately stand out as manifestations of culture because of their attributes as both denotative as well as connotative elements, used to evoke certain ideas in the mind of the speakers/readers familiar with their referents. From the list of procedures presented by Newmark (ibid.:69-93), he selects the following for the translation of culture (ibid.:103):

 Transference: Transfer a word from SL to TL. E.g. Malawi – Malawi

 Naturalization: Term is adapted to the pronunciation or morphology of

the TL. E.g. Edimburgh – Edimbourgh

 Cultural equivalent: Use a cultural concept from TL which approximates to the cultural concept used in SL. Baccalauréat – “A” level

 Neutralisation (functional or descriptive equivalent): Explaining with words the function or external characteristics of the cultural element ‘requires the use of a culture-free word’ (ibid.:83). Baccalauréat – French secondary school leaving exam’

 Literal translation: word-for-word translation Hall – sala

 Label: provisional translation (generally applies to new coined terms).

Heritage language – langue d’héritage

 Accepted standard translation: Use the recognised translation already in existence in the TL. E.g. Rechtsstaat – constitutional state

 Deletion: omission of redundant or less important elements

 Couplet: Combine two procedures

 Paraphrase, gloss, notes, etc.: Additional information added either to the main body of the text or as a footnote, glossary, etc. E.g. ‘at Handley’s’ –

‘dans le bar Handley’

 Classifier: Use of a generic term with the aim of making the term

clearer for the TL reader. E.g. Speyer – the city Speyer, in West Germany

Newmark concludes his exploration of translation and culture stressing that ‘more than in any other translation problems’ the translation of cultural words depends mainly and primarily ‘on the readership (…) and on the setting’ (ibid.:102).

Leppihalme (1997) (4.3.4) introduces a similar point of view, showing how, when the target readers are not taken into account, the cultural element sometimes fails to be transferred to the TL. Both ideas will be applied to the map of procedures described in Chapter 8 and used with the names in BRIDGET.

4.3.2 Hermans (1988)

Working at the same time as Newmark, Hermans (1988) addresses the problem of the translation of proper names from a different perspective, setting out to show how the study of the procedures employed with proper names in a translated text can shed light on the norms governing the act of translation. Hermans hypothesises that:

as far as the descriptive study of translation is concerned, proper names constitute a privileged object of enquiry. They stand out among other lexical items and can be isolated with relative ease […] the translational norms underlying a target text as a whole can in essence be inferred from an examination of proper names in that text. (ibid.:14)

Following on the tendencies emerging at the time (Chapter 1) which aspired to making Translation Studies an empirical science (Toury, 1995) , Hermans explored his hypothesis by examining proper names in the actual translations – one English and one German - of the novel De Witte by Erness Claes. In addition, he also looks at the proper names in eight different translations of the novel Max Havellar by

Multatuli.

In his study Hermans encounters several possible techniques for translating proper names. He classifies them as follows:

 Copy: the name is reproduced in the TT exactly as it was in the

 Transcription: name is transliterated or adapted on the level of

spelling, phonology, etc.

 Substitution: the ST name is replaced in the TT for any given

name in the ST (with a formally unrelated name)

 Translation: when a name in a ST is enmeshed in the lexicon of

that language and has acquired ‘meaning’ it can be translated.

(ibid.:13)

Hermans acknowledges the possibility of the combination of several of these modes and points out some alternatives:

 Non-translation: deletion of a ST proper name in the TT

 Replacement of a proper name by a common noun

 Insertion of a proper name in the TT where there is none in the

 Replacement of a ST common noun by a proper noun in the TT.

(ibid.:14)

In considering the various ways in which proper names are rendered in the different versions of these novels, Hermans illustrates a clear relationship between the translational norms in operation and the treatment of proper names. This brief work highlights the ‘great force’ that proper names have in a text and the even ‘greater force’ of proper names ‘in literary texts’ (1988:13) and show ‘the sometimes bewildering range of options and solutions which is not only theoretically available to translators but also used by them in practice’ (ibid:24).

4.3.3 Franco Aixelá (2000)

In an attempt to categorise this range of options, Franco Aixelá (2000) sets to explore the translation, from English into Spanish, of proper names within seven text types including different literary genres, essays and journalistic texts (already introduced in 3.3). He aims to account for all the procedures used by translators and to set them in a progressive scale measuring the degree of cultural transformation to which the names are subjected; that is whether they are kept in their original form or adapted to the target language culture. Franco Aixelá encounters a total of 12 procedures, which he divides into two blocks of six. The first block includes the procedures which show a higher degree of ‘conservation’ of the source culture. In the second block, called ‘substitution’, the procedures neutralize or omit the foreignness of the names or their exotic aspect as elements belonging to a different culture or environment, bringing them closer to the TL culture or making them culturally plain, not part of any specific culture.

Franco Aixelá’s map consists of the following procedures:

Conservation:

 Repetition: the proper name is reproduced in the target language with

the same form as in the source language

 Orthographical adaptation: changes in the spelling of the name but the

name still feels ‘foreign’

 Terminological adaptation: formal change of the name, when it has an

established form in the target language

 Linguistic translation: transfer of the semantic content of the name but

with the name still belonging to the source culture

 Extratextual Gloss: use of any of the previous strategies but with an

explanation or comment added by the translator outside the main body of

the text, e.g. in a footnote, a glossary, etc.

 Intratextual Gloss: an explanation is included in the main body of the

text.

Substitution:

 Limited neutralization: substitution of a proper name for a different

referent but still exotic or foreign in the target culture

 Absolute neutralization: the original proper name is transformed into a

cultural referent which cannot be placed in any society in particular

 Naturalization: substitution of a proper name for another which

specifically belongs to the target culture

 Ideological adaptation: when a name would be unacceptable in the

target language it is then replaced by another entity which would be more

ideologically acceptable for the readers

 Omission: deletion of a proper name from the TT

 Autonomous creation: introduction of a proper name.

The division of the procedures into two blocks in Franco Aixelá’s work shows a certain degree of gradation and hints at a cline with a natural progression between the procedures. That is, it can be argued that the use of a naturalization for instance, (replacing the SL name with another name from the TL culture) is a move further away from the ST than the use of a repetition where the SL name is maintained. Similarly, the different procedures within each block seem to represent different levels of adaptation to the new culture. This point will be addressed again in section 8.3 when the possibility of placing the procedures on a continuum is discussed in detail. It seems clear, nonetheless, that the Substitution and

Conservation blocks in Franco Aixelá’s model seem to correspond to Toury’s concepts of ‘acceptability’ and ‘adequacy’ (1995:56-9), each procedure being assigned to one block or the other depending on their level of manipulation of the SL term, and how this affects the overall tendency of the text. However, Franco Aixelá does not go into any detail regarding a more specific location for the procedures within each individual block. His main focus is on the general effect of the

85 procedures in the TL culture, rather than the detailed effect of each individual procedure in the translated text.

From this point of view, Franco Aixelá concludes in his study that in order to ensure that the TT proves valid from a communicative point of view, and thus admitted into the target culture, both from a commercial and a sociocultural point of view, the names are and should be adapted to that culture. He touches upon the Skopos theory, and the role that translated texts play in the TL culture and concludes that only by bringing the names closer to the TL audience, can the translations enter and thus be successful commercially in, the new culture.

4.3.4 Leppihalme (1997)

Whereas Franco Aixelá explores the way in which proper names adapt to the TL culture and the familiar universe of the TL readers, Ritva Leppihalme (1997) takes a different approach to exploring culture, exploring ‘what the words mean in a particular situational and cultural context’ (ibid.:viii). Leppihalme focuses on culture- bound elements which ‘are expected to convey a meaning that goes beyond the mere words used’ (ibid.). She studied ‘700 English-language allusions collected from

21 fictional and 200 non-fictional British and American texts’ (ibid.:5), from the twentieth century, and their translations into Finnish. Proper names are one of the culture-bound elements that she explores, together with key-phrases, clichés and proverbs. She takes into account both what she considers to be ‘potential’ (ibid.) procedures (which she terms ‘strategies’ (ibid.)) as well as ‘those actually adopted’

(ibid.) by the translators (that is, the actual solutions adopted). She takes her study a step further by also considering reader response, asking a group of 80 Finnish readers to analyse the translated texts to point out possible comprehensional problems or difficulties in understanding the meaning of any elements in the TTs.

Leppihalme’s study shows (ibid.:142) that readers’ comprehension of the TT was made easier when the translator had made use of procedures involving more changes or manipulation of the text, whereas the use of more conservative procedures, such as for example literal translations, would create more problems of understanding.

With regards to proper names, Leppihalme (1997:78-135) observes that they can be retained in their original form, modified or omitted, presenting the following possible procedures:

1- Retention of a name (either unchanged or in its conventional TL

form):

1a – use the name as such

1b – use the name, adding guidance

1c – use the name adding a detailed explanation, for example a

footnote

2 – Replacement of name by another

2a - replace the name by another SL name

2b – replace the name by a TL name

3 – Omission of name

3a - omit the name but transfer the sense by other means

3b - omit the name and the allusion all together.

The greatest value of this empirical work is to highlight the importance of a cultural analysis of the terms in their context as well as of taking into account all members involved in the translation process, translators and readers, when attempting to transfer the meaning of a culture-bound expression.

4.3.5 Marco (2004)

This chapter has so far shown how Newmark (1988) Hermans (1988) and Franco

Aixelá (2000) contributed to decipher the actual procedures used by translators for culture specific elements and how Leppihalme’s (1997) approach puts these procedures in the context of a cultural analysis. Following on from them, the novelty of Josep Marco’s (2004) classification of techniques for the translation of cultural referents lies in placing the techniques on a continuum (Figure 1). The extremes of

Marco’s continuum represent translator ‘intervention levels’ (ibid:136) – from less intervention, and thus more distance to the TL reader, to more intervention, and consequently closer to TL readers.

Figure 4.2: Marco’s continuum for translation techniques (ibid:138)

Marco (ibid:133) sees a continuum as the most operational way of representing the progression between translation techniques and remarks how taxonomies from other authors; such as for example that of Vinay and Darbelent, show a marked gradation.

This point has already been illustrated here (4.3 above) with regards to Franco

Aixelá’s (2000) two blocks, which also show a natural progression between the procedures. Marco goes so far as to link translator intervention to the relationship of the individual proper name to the target readers, but this connection will be questioned later in this study (section 8.3).

Marco identifies the following techniques:

 Borrowing: keeping a foreign word or expression without any

changes

 Literal translation: word by word translation

 Neutralization: offering an explanation of the function or

external characteristics of the cultural referent

 Amplification/compression: adding or omitting information

 Cultural adaptation: a source language cultural referent is

substituted by another source language culture referent but more

familiar for the target audience

 Intercultural adaptation: using a target language referent which

is equivalent to the source language cultural referent

Marco furthermore identifies three additional techniques: fixed equivalent, omission, creation. He leaves these techniques outside the continuum as they prove ‘too difficult to place’ (ibid:137):

 Coined equivalent: using an equivalent term or expression in the

target language which is recognised and accepted (for example, in the

dictionary) in the target language culture

 Omission: Omission of elements which are considered

redundant or not important

 Creation: insertion of a cultural referent in the TT where one did

not exist in the ST.

Marco’s model is applied practically by García Marín (2010) in her study of the translation from English to Spanish of legal terminology used in modern fictional texts (see also 6.2). García Marín successfully uses Marco’s model to identify and classify the legal terms extracted from a corpus of three novels and their

90 translations. She, however, acknowledges some gaps in Marco’s model and argues the need to add a further set of procedures, namely synonym and adaptation.

All in all, Marco’s greatest value lies in taking the initiative to set the procedures along a continuum as a way to illustrate the gradation between them. This point will be replicated with our model (Chapter 8) and it will also form an essential part of the data analysis stage (Chapter 9).

4.4 Conclusions

This chapter summarizes the key points and achievements of the works of Newmark

(1988), Hermans (1988), Franco Aixelá (2000), Leppihalme (1997) and Marco

(2004). These works approach the study of culture (or more specifically of proper names) from a procedural point of view and as such they all contribute to the map of translation procedures used in the present project. This map will be presented again in Chapter 8, with the aim of highlighting those particular aspects from each theorist that are to be incorporated into our propsed model and new map of procedures.

Before moving onto the methodology for analysis, it is first important to define and characterize the genre that forms the corpus of novels analysed in this study, the

Chick Lit genre. Chapter 5 presents the main characteristics of the genre describing its main attributes and reviewing the main theoretical works written on the subject,

91 in an attempt to contextualize the practical work undertaken and covered in the subsequent chapters.

Chapter 5. Chick Lit

“Names! There are so many. Each one cleverer or more far-fetched than the last” Bielski, N.: The Year is ’42 (p.187)

5.1 Introduction

This chapter aims to offer an indication of the main characteristics of the Chick Lit genre as a way of contextualizing the research. At the same time it sets out the main reasons for the selection of Chick Lit as the source of data to conduct this investigation into translated proper names. With this in mind, a summary of the novels included in the corpus is then offered in order to introduce the characters and the main plots in them. As advanced in previous Chapters (mainly Introduction and

Chapter 2), proper names are often used in these novels as connotative elements to offer additional information aimed at native readers. The context and the single referent behind each individual use of a name are thus paramount to fully comprehending the meaning of the names. For this reason, a summary of the main points of the novels narratives is extremely important.

5.2 Definition of a new genre

The characteristic most typically used to define the Chick Lit genre is that it is written for women by women (see for instance, Gormley, 2009, Harzewski, 2011,

Montoro, 2012). There is general consensus (Ferris and Young, 2006, Gill and

Herdieckerhoff, 2006, Gormley, 2009, Modleski, 2008, Yardley, 2006) in signalling

Fielding’s Bridget Jones Diary, published in 1996, as the first novel of the genre.

Chris Mazza (2006) considers herself the first to coin the term ‘Chick Lit’ in the mid- nineties (Mazza and DeShell, 1995) to ironically refer to postfeminist women’s fiction which combined disinterest in political issues with the most popular female stereotypes; the term gained popularity together with Fielding’s novel and entered general usage to refer to this literary phenomenon. To a certain extent, it is the link of the Chick Lit novels to postfeminism which has generated the majority of the academic debates around the genre, as indicated in more detail below.

From a literary point of view, Chick Lit novels follow certain lines and plots which can be generally summarized as follows: the protagonist, typically a middle-class white woman, struggles to find a balance in her life, always in search for the perfect partner, the perfect house, the perfect job and the perfect outfit. These novels move in the reality of modern life for many urban women, and attempt to touch upon more serious subjects from a light-hearted and humorous position. One of the most exhaustive and wide-ranging definitions of the genre can be found in Mlynowsky &

Farin A Girls’s Guide to Writing Chick Lit:

Chick Lit is often upbeat, always funny fiction about contemporary female characters and their everyday struggles with work, home friendship, family or love. It’s about women growing up and figuring out who they are and what they need, versus what they think they want It’s about observing life and finding the humour in a variety of situations, exchanges and people. It’s about coming of age (no matter how old the woman is – chick Lit heroines can be anywhere from teenaged to beyond middle-aged). It’s generally written by women for women. It’s honest, it reflects women’s lives today – their hopes and dreams as well as their trials and tribulations – and, well, it’s hugely popular. (Mlynowski and Farin, 2006:10)

Thus, the most noticeable trait of the genre is the representation of modern women’s life. From this standpoint, the origins of these novels are often traced back to fiction by women during the nineteenth century, most importantly to Austen and the Brontë sisters (Harzewski, 2006, Wells, 2006), (see 5.3.3 for a direct analogy of

Pride and Prejudice and Bridget Jones’s Diary). In a previous work, (Cantora, 2010) I highlighted the way in which Chick Lit novels, through the modern lives of their protagonists, aim to recreate the issues, problems and concerns faced by women nowadays. Fielding herself confesses in an interview that ‘it is good to be able to represent women as they actually are in the age in which you are living’ (Ezard,

2001). It is possibly this point, when readers can identify with characters, where the commercial value, and most importantly success, of these novels lies. Montoro

(2012) argues that ‘the genre needs to be understood in its context of production and reception’ (ibid.:1). I join voices with her to emphasise, as already sketched out in the Introduction (section 1), that the way in which Chick Lit authors manage to recreate a world where readers can directly relate to characters is deeply rooted to the culture where these novels originate. This is done by means of setting these novels in a contemporary location, familiar to the intended readers (in most cases

London, certainly in the three novels in BRIDGET), as well as through the products that they consume and the places they go to purchase them. All these are based in reality, and exist outside the world of the novel, and in a way they evoke a sense of verisimilitude and familiarity where both characters and readers can feel at ease. For instance, when Bridget Jones meets her friends they do so in Café Rouge; this particularly appeals to British women. Similarly, Amanda taking the District Line to work or Rebecca Bloomwood longing to be invited for dinner at The Ivy, are facts

95 that contribute to making these novels seem so familiar and realistic to their intended readers. By assimilation, this specific socio-cultural backdrop makes these novels so rich in proper names, and this in turn makes Chick Lit an ideal genre to base this research in. Thus, what Montoro states about the whole genre can more specifically be applied to the proper names in these novels: they are taken as a representation of culture and as such they have to be understood in the context where they occur but also taking into consideration the background where they originate, as well as the ‘intended meaning in the mind of the author framing the message’ (Leech, 1981:14). This point, then, validates the choice of genre and in addition represents the pioneering feature of this research. This exploration of proper names from such a modern and culturally embedded genre affords a new perspective to both the genre and the subject of the translation of proper names.

Other typical features of a Chick Lit novel include a stereotypical characterization of the genre’s heroines often seen as ‘women seeking fulfilment in a romance- consumer-comedic vein’ (Knowles, 2004:3). By assimilation to Bridget Jones, these women are often also seen by critics as trivial, consumerist and obsessed with losing weight. But these traits cannot be attributed to every character. They all have their struggles and personal issues, but not all of these are consistently related to weight or financial problems; although admittedly, there is always space for romance in their lives. Another successful Chick Lit author, Joanna Trollope, expresses her concern with the disdain given to the permanent presence of love plots in these novels: ‘There's a squeamishness in the literary world about love, but love is the great topic […]. I think it produces a fear amongst some people in the industry that

96 if they have to write about their feelings, they'll confess too much’. (Jones, 2008).

Besides, as Harzewsrki (2006) points out ‘contemporary Chick Lit doesn’t always end in marriage and the heroine’s wider social life plays just as large a role in her life as heterosexual coupling’ (Harzewsrki, 2006.:38). Notwithstanding the extent of the heroines’ obsessions with more or less trivial matters, the dominant narrative strategy in the genre is its ‘use of a first person female narrator’ (Ibid.). Therefore, another specific feature of these novels is that they are almost always recounted in a first person, from the point of view of their protagonists.

Another trait typical of Chick Lit novels is represented by the ‘clearly marked jacket designs’ (Gill & Herdieckerhaff, 2006:488). Montoro (2012) devotes an entire monograph to studying this issue, both from a literary and a stylistic perspective.

She argues that it is the stylistics of the genre that more strongly have contributed to ‘how it has set itself apart from related novelistic forms’ (Ibid.:1). It cannot be denied that these novels are consistently presented in a similar linguistic realisation, which immediately stands out and attracts the attention of its intended readers.

These features include pink or pastel coloured covers with pictures of shoes, handbags or cosmetics and twisted fonts. In contrast, these characteristic features are often also seen as a detriment to the genre as it deters potential new readers, who would prefer not to be seen with such a book in their hands. Author Louise

Renninson, for instance, complains about the effect that the pink colours have for her novels: ‘boys will never pick up a book that bellows its girly credentials outright, even if they secretly harbour some curiosity about what it is girls read, want and incessantly giggle about’ (Russell, 2010) .

Despite the success of these novels the genre is often condemned for its lack of linguistic or literary value, and seen as trivial and offering nothing to their readers.

Most famously, writer Beryl Bainbridge (2001) rejected the genre as ‘froth’ at the same time that Doris Lessing (2001) lamented that women waste their precious time reading such novels. Taking personally Bainbridge’s accusation, Chick Lit author

Jenny Colgan (writer of Amanda’s Wedding, included in BRIDGET) defends the genre’s right to entertain via a humorous take on life-matters: ‘We know what foie- grass is, Beryl, but sometimes we just want Hulla Hoops’ (Colgan, 2001). Colgan stresses the value of the genre as a comedy genre. In the same line, Fielding counteracts some of the criticisms of Bridget Jones as she says she did not intend her to be taken seriously (Ezard, 2001).

In addition to these criticisms by fellow writers, the more scholarly attacks come from the postfeminist discipline, where these novels are often grounded. The value of the genre as empowering to women is often seen as a disgrace to feminism, for example with such an explicit link between women’s independence and liberation to consumerism (Wilson, 2012). Furthermore, feminist pop culture critic Maher (2007) wonders whether aspiring to obtain the latest designer handbag being portrayed as more important than pay equality can be praised as positive. She states how she is

‘suspicious of representational “empowerment” discovered via female characters who can run in high heels, fire a gun and bed the boss’ (Ibid.:195). Maher laments the complacency of authors and characters of Chick Lit novels with the real situation of many women nowadays, still being paid less than men, for instance, and would like to see the heroines positioning themselves more on the grounds of more

98 feminist concerns rather than ‘lifestyle’ (Ibid.) concerns. However, I would argue that these novels never aspire to become the manifestos of the postfeminist movement, but rather simply to offer a representation of the lives of many women today, without necessarily aiming to change them or denounce them as complacent.

And this is perhaps where the success of the novels lies. Readers can distract themselves from their own problems by delving into the similar problems faced by characters to whom they can relate, without having to be made to feel guilty that they worry more about how to obtain the new Jimmy Choo shoes than about world peace. This is not the genre for those issues. This is not what the Chick Lit readers are ultimately looking for. And the Chick Lit novelists have understood that very well, as can be attested by the huge sales figures for these books. For instance, Chick Lit author Sophie Kinsella ‘has the biggest backlist sales out of any female author in the

UK’ (Jones, 2008) and figures for Chick Lit sales in the U.S in 2005 were predicted around $137 million (Ferris and Young, 2006, Gormley, 2009, Montoro, 2012).

The genre is also a translation phenomenon: these novels are often translated to several languages shortly after their publication date in the countries where they originate (most commonly Britain and the US), a factor which should make them relevant enough for the TS discipline, regardless of the arguments about their

(commonly questioned) value as a genre.

5.3 Chick Lit in Spain and Italy

Chick lit novels in Spain and Italy are generally classed as a subcategory of the romance novel and initially entered these markets through translations. Well known

99 authors such as Sophie Kinsella or Marian Keyes are a reference in the Spanish and

Italian popular fiction markets and often all their works are translated short after their publication dates in their original countries and see many reprints.

The three novels that make BRIDGET, Amanda’s Wedding, My Life on a Plate and

Bridget Jones’s Diary all have several reprints in both markets, which can also be taken as a reinforcement of the popularity of the genre. For instance, Bridget Jones’s

Diary was first printed in Italy in 1998, only one year after the publication in the UK; in 2002 it reached the 12th edition. Similarly, My Life on a Plate was published in

2001, 2003 and 2006, this last edition reprinted 5 times. On its part, in the Spanish market the popularity of Bridget Jones’s Diary is marked by the 12 different editions published between 1998 and 2005. My Life on a Plate has 4 editions. All the novels can be found in both hardback and paperback formats and the novels follow the trend for the genre set in the UK and USA, with pastel-coloured covers. However, the Italian editions occasionally also show more plain pictures rather than just drawings of handbags or high heels, so common of these novels in the UK.

These novels are generally received in the TL literary markets as a representation of the lives of British or American women, and taken as a description of the fashions and ways of life in London or New York, where the novels are more frequently set.

Following the popularity of these authors, novels originally written in Spanish showing the traits (described above) of the genre started to spark from the late

2000s onwards (Blanco, 2009). The popularity lead, for instance, the Spanish

100 publishing company Planeta to the creation in 2009 of an imprint, Esencias, devoted exclusively to Chick Lit novels.

The first novel published by Esencias, Aceptamos marido como animal de compañía written by Miriam Lavilla Muñoz became one of the bestselling books written in

Spanish for that year.

Rebeca Reus is another renowned author in the genre. Reus first two novels are based on the life of Sabrina, a twenty-something careless woman who suddenly decides to turn her life around and become a successful business woman. On similar lines to Bridget Jones, Sabrina’s personality usually gets her into trouble in comic situations and tribulations.

In the Italian market Chick Lit is a popular genre better known through translation, generally with more British or American authors translated into Italian and more reprints of the novels than in the Spanish market. However, Chick Lit novels in

Italian have been written by authors such as Alessandra Casella (Un anno di Gloria),

Tiziana Merani (Devo comprare un mastino) or Chiara Santoianni (Il diario di Lara).

This brief analysis of the reception and creation of this genre in the corresponding markets of the TLs of this study evidence the relevance of translation for the genre, in addition it also shows the extent of the popularity of the best known British Chick

Lit authors and novels.

101

5.4 Summary of selected books

5.4.1 Introduction

Having set the characteristics of the genre and the impact of these novels in the

Spanish and Italian literary markets, this section is now devoted specifically to the three Chick Lit novels which form BRIDGET. As general background to the research the plots of each book and certain key points in them are included below, under individual headings, one per book. Bibliographic details about the translations are also given. This section aims to act as contextualization for the analysis of proper names and the translation methods employed. Chapter 6 (section 6.3.1) offers an account of the selection process which took place in order to choose the novels to be included in the corpus.

5.4.2 Amanda’s Wedding

Jenny Colgan’s first novel, Amanda’s Wedding, was first published by HarperCollins in 2000. It quickly became a bestseller and has been reprinted several times.

Amanda’s Wedding tells the story of Melanie and her friends as they try to stop the wedding of their evil friend Amanda to the lovely Fraser McConnald, Melanie’s first crush during their university days. The group of friends believe that Amanda is only interested in Fraser due to the fact that he has just inherited his grandfather’s laird title. Fraser’s brother, Gus, is determined to show Fraser who Amanda really is.

While attempting to ruin the wedding, Melanie also has to deal with her disastrous

102 life, as she is demoted at work, she believes that her flatmate fights against bulimia, and her best friend starts a secret relationship with her boyfriend, to name but a few of her tribulations. The novel finishes on the actual wedding day with Amanda exposing her real character not only to Fraser but to all the many attendees as well.

Fraser then runs away into the arms of Melanie.

Amanda’s Wedding follows the typical pattern of a Chick Lit novel, recounting the struggles of a young professional woman to make sense of her life. The action is set in central London, although the novel also has a strong Scottish feel, as the main characters Fraser and his brother are Scottish.

With regard to the use of proper names, there are a lot of references to popular culture, such as Mystic Meg, Disney characters or EastEnders, as well as to popular places fashionable in London in the early 2000s, for example The Asprey or the

Princess Louise Pub in Holborn. Similarly, there are also references to Scottish society such as the Caledonian Ball, for instance. The novel has a comic tone to it and is full of stereotypes jokes and twists.

Amanda’s Wedding was translated into Spanish by María José Chávez Abad and

Armando Figeroa Rojas with the title La boda de Amanda. The novel was published in 2003 by Suma de Letras, S.L.

Non sposate quella donna!, the Italian translation, was undertaken by Valentina

Danielle and was published in 2001 by Salani Editore s.r.l.

103

5.4.3 My Life on a Plate

My Life on a Plate, published in 2000 by Penguin books. The author, India Knight, is a well-known British journalist with columns in several newspapers and magazines.

My Life on a Plate’s plot to some extent differs from what is expected from a more typical Chick Lit novel. Instead of a twenty something, the main character is in her mid-thirties, and rather than struggling to get her adult life started she is going through a crisis in her already established relationship and life. Although Clara has it all, she does not seem to be happy, but of course she cannot identify what is lacking: her husband is attentive, her young children perfect, her journalism job ideal,… In the end, after her husband leaves her, she is forced to face her fears and make a fresh start, actually aiming to achieve her real dreams.

Although the plot and main characters’ background divert from a typical Chick Lit setting, the novel remains undeniably within the genre. It attempts to represent modern society and ways of life, and recreates a scenario where modern women in their thirties can see themselves reflected. The upbeat ending brings hope and light.

In order to represent these ideals and to describe Clara’s life, Knight uses many references to modern London and modern middle class society, for instance dining at The Ivy, going to watch a performance of modern dance at Sadlers Wells, or spending a romantic weekend in Paris. The real references are numerous, and using the children as props, also include references to popular characters and TV programmes such as Blue Peter or Jar Jar Binks.

104

The Spanish translation, Mi vida es casi perfecta, was undertaken by Eduardo G.

Murillo and was also published in 2000 by Debolsillo.

The Italian translation, La mia vita su un piatto, was translated by A. Dallaglio.

Published in 2003 by Feltrinelli Editori.

5.4.4 The Diary of Bridget Jones

Helen Fielding first novel, The Diary of Bridget Jones, is the most famous novel in the genre, and as already stated, it is often considered the origin of Chick Lit, sparking with its success the publication of other novels with similar features.

The novel follows the life of the, now well-known, character, Bridget, from one New

Year’s celebration to the next. Through the different calendar months, the reader learns and hears of Bridget’s troubles at work, with love, with her parents, her struggles against the constant references to her still being single, with not losing weight… all the stereotypes which have since come to describe the genre.

Helen Fielding openly told BBC News of her admiration for Jane Austen and does not hide the fact that she took inspiration for the plot of Bridget Jones from Pride and

Prejudice (Fielding, 2013). In fact, the similarities between the two novels are striking; both have an unconventional main character who does not perform to the types of their period, both develop feelings for a not so desirable man and end up being rescued by the one they originally thought of as deceiving and rude. Both

105 include cases of elopement and embarrassment for their families; both conclude with a happy ending.

In her attempt to update this classic novel, Fielding describes modern society, the typical places visited, most often Cafe Rouge, the products consumed, Chardonnay

Wine or Pied-a-Terre shoes, for instance. On the other hand, and following the traits of the genre, the story is told in a humorous tone, with numerous jokes and funny passages.

The Spanish translation, El diario de Bridget Jones, was done by Néstor Busquets, and first published in 1997 by Lumen Editores.

The Italian version, Il Diario di Bridget Jones, was published by Sonzogno Editore in

1998 and translated by Olivia Crosio.

5.5 Conclusions

This Chapter contextualizes the genre that is the base of this research. Chick Lit novels are hugely popular and generating a lot of translation work and their direct links, as seen above, to popular culture make them a valid area of study. In addition, and more pertinent for this particular project, as highlighted in section 5.2, these modern novels are deeply embedded in the British culture where they originate and as such are very rich in proper names, used by authors as the main elements to

106 characterize the modern reality they want to portray. From this standpoint, this genre stands out as the best source of data for our corpus (see section 5.3).

This Chapter concludes the theoretical aspects of this thesis. The Chapters that follow are devoted to the empirical investigation which took place from the bases of this theoretical background; starting in Chapter 6 with the corpus methodology employed in the design, creation and interrogation of BRIDGET.

107

Chapter 6. Corpus Methodology: Data Acquisition

“Of course animals also make warning calls when danger threatens, but they don’t have names for things as human beings do”. Gombrich, E.H.: A Little History of the World (p.7)

6.1 Introduction

Chapter 1 (Section 1.4.1) briefly introduced the way in which Corpus Linguistics has made its way into Translation Studies as a methodology and offering the tools to study particular features of translated text (called translation universals by Baker,

1993) as well as the processes that lead from ST to TT (as envisaged by the DTS approach introduced by Toury, 1995). Laviosa (2002 & 2011) gives an insight into the extent to which this new approach has shaped the TS discipline during the last decades. She stresses its role as a new paradigm to offer empirical validity to the discipline (Laviosa, 2002:119) in as much as equipping researchers with ‘tools and techniques that can truly improve the quality and efficiency of their work’ (ibid.:1).

However, there are also calls for ‘systematic description’ (Oakes and Ji, 2012:1) of these tools and techniques in order to construct and test ‘theoretical models for literary translations’ (Ibid.). In the present chapter I will address these points from a practical and methodological point of view. Linking an overview of the literature around corpus compilation and interrogation and descriptions of successful CBTS projects with the specific issues faced when compiling and interrogating BRIDGET.

By describing in detail the processes followed in this particular research, the tools employed and their success in offering the desired results, the project contributes to the expansion of the field, with methods, tools and processes that can be verified,

108 tested and replicated in further research. In this particular project, corpus tools are taken as the means to facilitate the study of the elements under investigation.

Saldanha (2009) explains that ‘corpus linguistics is not a linguistic theory but a methodology’ (Saldanha, 2009); this point of view reflects the main reasons for using corpus tools in this project: primarily the intention is to automate the identification of the proper names in the English texts as well as of their corresponding translations in the Spanish and Italian versions. Secondly, having an electronic corpus of the novels facilitates the quick access to the data, and its context, and renders the analysis more efficient, making it possible to work with more data than it would otherwise be feasible to manipulate manually.

The chapter starts with an overview in Section 6.2 of the literature around corpus compilation and interrogation and a description of corpora and CBTS projects that embrace a similar methodology to the one described here. The main terms and concepts in the field are explained and defined; those relevant to this study (parallel corpus, concordance, retrieval systems, for instance) will be covered in detail in

Section 6.3 linking them directly to the particular issues faced in the compilation and interrogation of BRIDGET as well as to the work and specific tools used in this research on translated proper names.

6.2 Corpus design, compilation and interrogation

A corpus is ‘a large collection of authentic texts’ (Bowker, 2002:9). McEnery and

Wilson (1996: 87), define a corpus as ‘a body of text which is carefully sampled to be maximally representative of a language or language variety’. Following Kruger et

109 al. (2011) a corpus in modern corpus-based translation studies research is by default

‘understood as a collection of texts in electronic form’ (ibid.:1), as it is the advances in computers and the development of specialized electronic tools for the analysis of texts that has empowered the rapid growth of the discipline.

Johansson (1998) describes a corpus as ‘a body of texts put together in a principled way and prepared for computer processing’ (ibid.:3) All these definitions stress different attributes of a corpus suggesting that there are several questions to consider when attempting to design an electronic corpus. Firstly, a corpus has to be created ‘according to a specific set of criteria’ (Bowker, 2002:9). Maeve Olohan stresses that ‘these criteria are based on the aim of the research, the research questions to be addressed and the hypotheses to be tested’ (Olohan, 2004:46).

Similarly, Kenny points out that ‘the design of a corpus, and the selection of individual texts for inclusion in the corpus, are determined principally by its envisaged purpose’ (Kenny, 2001:106). Munday (1998) links this issue directly to the research questions to be answered, the features to be investigated and, more importantly, the tools available to interrogate the corpus and produce answers to the points that are being addressed.

These aspects have a direct impact, of course, on the type of corpus to be created.

There are many types of corpora, which can be used for different kinds of purposes

(cf. Kennedy, 1998). Bernardini (2003) considers key questions of corpus types stressing that ‘the terminology in this area is not consistent’ (ibid.:5). In an attempt to offer a framework in which to describe each type of corpus Laviosa (2002:34-38)

110

presents a typology in four levels, from the most general features to the more

specific. Table 6.1 below represents Laviosa’s Typology:

Laviosa Hierarchical Typology for Corpus Level 1 Level 2 Level 3 Level 4

Corpus Types FULL-TEXT Multilingual Bilingual Monolingual SINGLE Corpus Single TRANSLATIONA Corpus Translational MON L O SL

SAMPLE COMPARABL NON- BI SL E TRANSLATIONA

MIXED PARALLEL Corpus Parallel Bilingual MONO- MULT DIRECTIONAL I SL

MONITOR BI-

DIRECTIONAL Types Corpus SYNCHRONIC Corpus Parallel Multilingual MONO-SL DIACHRONIC BI-SL

Types Corpus GENERAL MULTI SL TERMINOLOGIC

Types Corpus MONOLINGUAL BILINGUAL MULTILINGUAL

Types Corpus WRITTEN SPOKEN MIXED Table 6.1: Representation of Laviosa’s (2002) Corpus Typology

An exhaustive categorization of corpora lies outside the scope of this chapter; it is

important, however, to consider certain types of corpora which relate directly to the

different features of BRIDGET and which would facilitate a better understanding and

description of the particular features included in this corpus. Taking Laviosa’s

hierarchical approach, the first and broadest considerations relate to the more

general features involved in the texts contained in the corpus: Regarding the

languages; a corpus is monolingual when it contains texts in one language,

111 bilingual when it contains texts in two languages or multilingual if it contains texts in several languages. If these texts represent spoken language, it is a spoken corpus, whereas if they represent written language then it is a written corpus; a corpus is mixed when it contains representations of both. Johansson’s definition of corpus presented above made special consideration of the fact that an electronic corpus is ‘prepared for computer processing’ (Johansson, 1998:3); from this point of view, if the texts have been treated a corpus is then annotated or tagged

(lexically, morphologically, grammatically, and so on); a corpus is raw if the texts are left untouched. From these parameters, BRIDGET (more details in 6.3.3) is a multilingual, written corpus initially left in raw format.

Following on with Laviosa’s hierarchic typology, at the second level there are parallel and comparable corpora. Whereas a parallel corpus is made of SL texts and their corresponding translations in the TLs (facilitating direct comparison of SL vs TL segments) a comparable corpus, on the other hand, is ideal in research trying to identify the specific patterns or traits of translated language, as Mona Baker (1993) pointed out when first exploring the possibilities of adopting a corpus methodology in translation research.

Examples of comparable corpora are general or reference corpora, such as BNC or the Bank of English (already introduced in section 1.4.1) which represent a language or variety as a whole.

112

The BNC comprises more than 100 million words representing both written and oral contemporary English. The compilation of the BNC took place during the period that

Laviosa (2011) describes as the ‘dawn of corpus-based translation studies’ (Ibid.:13) between 1993 and 1995 (see 1.4.1). The BNC offered the means to test and reference the findings from translation research and facilitated the exploration of recurrent features of translated text (not present, or present with a higher –or lower- frequency in these corpora than in corpora formed of translated texts). Thus, reference corpora act as a guide against which to compare the results obtained from the analysis of a parallel corpus – or of any other type of corpus. Monolingual corpora such as the BNC, or the CREA (Corpus de Referencia de la Real Academia) in Spain, are noteworthy reference corpora; if the researcher is trying to describe the specific features of translation, comparing the findings highlighted by the examination of patterns emerging from a parallel or comparable corpus with language directly written in the SL, as opposed to translated language, they provide an instrumental way of validating or discarding the results.

A Parallel corpus is a corpus that holds the same texts in more than one language, i.e. STs in one language and their translations into one or more languages. In contrastive studies a parallel corpus is sometimes understood as original texts in the same domain, genre or text types in language A and B (Laviosa, 2002:37). This study embraces the first definition: texts in language A and their translations into language B, on the grounds of representing the contents of BRIDGET as well as of being the most widespread definition in CBTS (McEnnery & Wilson, 1996; Laviosa,

2002; Olohan 2004).

113

‘Parallel corpora can be unidirectional, i.e. STs in language A and TTs in language B, or bidirectional, i.e. STs in language A and translation in language B, and STs in language B and their translations in language A’ (Olohan, 2004:24).

Hence at this level, BRIDGET is a multilingual, unidirectional parallel corpus formed by original British Chick Lit novels and their translations into Spanish and Italian.

In translation studies during ‘the dawn of corpus-based translation studies’ (Laviosa,

2011:13), the most remarkable work of this type was undertaken by Stig Johanson

(2003) in the design and compilation of the English-Norwegian bidirectional parallel corpus (ENCP). This corpus contains original texts and their translations, bidirectional

English-Norwegian and Norwegian-English. It includes both fictional and non-fictional texts and comprises a total of 2.6 million words.

From the mid-nineties onwards, the number of corpora being created and the number of studies taking a corpus-based approach started to grow exponentially.

The initial insights into the use of corpora for investigating specific features of translated language were taken even further, uniting corpus to Toury’s methodology and demonstrating the value of using computerized tools also in research into the mechanisms that move translation.

One prominent example of research on these grounds is the work undertaken by

Dorothy Kenny (2001). Kenny built her own parallel corpus of original German texts and their English translations, GEPCOLT, with an aim to study creative word forms

114 and collocations in the texts. By using different corpus tools and processes she succeeds in tracking and retrieving each author’s individual lexical creativity. For instance, she uses frequency lists to isolate every single word that occurs only once in the German (STs) sub-corpus of GEPCOLT; assuming that if a word occurs only once it is potentially an example of the author’s own creativity. She then uses data from reference corpora, such as the BNC for the English translations and the

Manheim Corpus for the German originals, as a control measure. If these unique words identified as signs of one particular author’s creativity do not appear with some degree of frequency in the reference corpora it can be assumed that they indeed belong to the author’s personal style. Comparing these creative points with their translations into English, she finds evidence of a tendency to a lexical normalization in the translated texts, although she states that ‘certain translators may be more inclined to normalize than others’ (Kenny, 2000:211).

Along similar lines, Charlotte Bosseaux (2007) sets to find the translators own voice and style at discourse level in translations of Virginia Woolf’s novels into French.

Bosseaux tracks traces of the translator’s own work and initiative in the TTs using corpus tools. In particular, she uses WordSmith’s (Scott, 1997) Wordlists with the

STs and Multiconcord for generating concordances in the TTs (more details in 6.3).

Bosseaux then combines the data extracted from these tools with Baker’s (2000) proposed methodology for identifying translator’s own style. With this study

Bosseaux ‘showed that it is possible to identify the “other voice” of translation and the style of an individual literary translator and that in addition to the translators’

115 strategies it is even possible to identify the kind of world that each translator has decided to recreate’ (Bosseaux, 2001:73).

Also specialising on translational stylistics, Saldanha (2005, 2011) argues that it is possible to identify the translators own stylistic preferences. Saldanha works with two corpora of Spanish and Portuguese texts and their translations into English.

These corpora collect some of the works translated into English by Margaret Jull

Costa and Peter Bush. ‘Each corpus consists of five STs, by five different authors and their translations’ (Saldanha, 2011:241). In total, they make over 650,000 tokens.

Saldanha, like Bosseaux, follows Baker’s (2000) methodology for the study of translator’s style, but combining elements such as parallel and comparable corpora in an attempt to identify the traits that can be attributed to the translator.

Specifically she looks for uses specific to the translator’s own style of culture-specific items, emphatic italics and of the connective ‘that’ after verbs of speech. Saldanha uses measurements such as normalized frequencies in order to assess whether the frequencies of use in the translations of the features under investigation differ from their frequencies in the original; in addition she contrasts this data with data extracted from a reference corpora (the bi-directional parallel corpus of Portuguese and English narrative COMPARA) in order ‘to determine to what extent’ (Saldanha,

2012:247 the findings represent distinctive uses of translators own preferences. She concludes that the analysed data points towards visible choices in the translations that are distinctive characteristics of the translator’s style. More interestingly, she combines her findings with meta-linguistical information collected externally to the body of the texts, such as bibliographical data on the publishing companies,

116 biographical details on the translators and personal communications and interviews with them. These leads her to affirm that ‘the most important factor in determining the translator’s choice is their different conceptualization of their role as intercultural mediators, in particular in relation to their readership’ (Ibid.:255). These ground- breaking findings call for the inclusion of additional ways of analysing translated texts; some perhaps going further than information that can be obtained solely from the body of texts that form the corpus.

Also on similar lines to the research reported here is the work of García Marín (2010) investigating the translation into Spanish of legal cultural referents in modern British legal fiction. García Marín presents a compelling comparison of the legal systems in both countries and uses a purpose-built corpus of original English legal novels and their translations into Spanish to explore the way in which translators bridge the gaps between both legal systems and cultures. Following the taxonomy for translation procedures presented by Molina & Hurtado (2002) as well as that of

Marco (2004) for translating legal terminology, she uses different statistical measurements to account for the procedures employed by the translators. She combines this quantitative analysis with a qualitative exploration of the more problematic terms (generally those that cover for terms in the ST culture that do not have a representation in the TT culture such as for instance hung jury). She concludes that translators tend towards a balanced position, using terms closer to the SL for those concepts that exit in both cultures and adapting the terms that do not represent concepts available in the TL culture.

117

On a much smaller scale, but still relevant for this study, is the research on the use of parallel corpora to infer translation procedures for culture-specific elements presented by Pearson (2003). In an attempt to demonstrate the benefits of introducing parallel corpora in translator training, Pearson reviews the procedures employed to translate, from English into French, names of different universities, academic institutions and researchers. The data is extracted from a small (356,000 tokens in total), purpose-built parallel corpus, of scientific papers published in US and their translations into French. Even at such a small scale, Pearson is able to infer different procedures used with these proper names; for example omission of the reference altogether stood out as a prominent procedure used, for instance, with 36 out of 102 instances of names of universities. With this brief review Pearson manages to demonstrate that ‘there may be many different answers to what might appear to some to be a simple question [the translation of culture-specific elements], and that study of a parallel corpus might reveal solutions that had not been imagined’ (Ibid.:23).

The research reported here follows similar methodologies to those used by Kenny

(2001) and Bosseaux (2007) in their projects. The process for creating a parallel corpus and for locating the elements of study in the corpus show similarities and this project will draw from findings and solutions presented to problems in those two works. However, the outcome of the projects is fundamentally different, as Kenny

(2001) and Bosseaux (2007) both set to find creative points in the text, the former on STs and the latter on the translations, whereas, for this research, corpus tools

118 will be applied, in the first instance solely to automatically identifying the names in

STs and in the second stage to find the names in the translations.

The process for analysing the data will be more on the lines of that employed by

García Marín (2010), selecting a model of translation procedures (Chapter 8) and applying it to the names extracted from the corpus. The potential of statistics and similar features of computerised programs, in lines of those used by Saldanha (2005

& 2011) will be used to generate results and to infer some findings from them, but not for the identification of specific features of translation, the ultimate aim of

Kenny’s, Bosseaux’s and Saldanha’s projects. The aim of this particular project on the translation of proper names is more similar to that of García Marín’s (2010); that is identifying and describing translation procedures employed by translators to shed light into the relationship between ST and TT segments, in the lines of Toury’s approach to DTS. Following Pearson’s (2003) a parallel corpus will be used here to

‘learn how information is conveyed, whether any information is lost, adapted or misrepresented in the process’ of translation (Ibid.:17).

The first step in the corpus-based projects described above as well as in this one, is the creation of the corpus which will serve as the source for the data. Once the type of research, the research questions and the type of corpus needed are established, the next big issue faced by corpus compilers is representativeness; that is a corpus has to be representative of the language under investigation. This is often shaped by external issues, such as copyright for example. When compiling a corpus where the texts to be included are under copyright laws, special consent to manipulate the texts has to be granted by the copyright holder. Identifying who holds the right to

119 one particular text can often be a complex task, and it is even more complicated when working with translated texts; this was a paramount problem encountered when attempting to contact the Spanish publishing companies for the texts selected for the inclusion in BRIDGET, section 6.3.1 refers back to this point in detail. These issues are often faced by researchers compiling a corpus (Hareide and Hofland,

2012:84, Saldanha 2011:441, Kenny, 2001:115). It is not until these issues have been overcome that the researcher can move onto compiling the corpus, that is obtaining the texts in electronic form. These difficulties are often seen as drawbacks for researchers considering compiling their own corpus and have a direct impact in the way the discipline has taken shape in recent years.

In research where the corpus has to be manually created, such as the above mentioned projects and the research reported here, once the criteria for the creation of the corpus have been determined, the texts to be included identified and copyright issues taken control of, the corpus compilation stage can take place. If the texts are not available in electronic format, this stage usually involves the use of a scanner for manual digitalization as well as the use of an Optical Character

Recognition Program (OCR) for the conversion of the images into texts. (Section

6.3.1 below has details of the exact processes followed in the creation of BRIDGET).

Regardless of whether scanning is required or not, proof reading checks are also required. In general, this is often a solitary and arduous task for the researcher.

Nonetheless, the end result, the compiled and ready-to-use corpus, is a very rewarding outcome.

120

Olohan explains that ‘once a corpus has been compiled or selected appropriate to a specific research question, the researcher has to decide what kind of data are required from the corpus, and in what form’ (Olohan, 2004:63). Text-processing tools, such as those included in WordSmith Tools suite developed by Mike Scott

(Scott, 1997), can be used to generate, for example, word lists and frequency lists

(used in projects such as Kenny’s (2001) and Bosseaux’s (2007) mentioned above), which list all the words in the corpus either in alphabetical order or in order of frequency. These can be used, for example, in research that is trying to establish the word frequency versus word significance in a given literary genre. This data can also be used to calculate the richness of the vocabulary, known as lexical variety, used in a given corpus. ‘The results can be presented in tokens, numbers of running words in the corpus, or types, actual distinct word forms; the higher the figure obtained by dividing the types by tokens, the higher the lexical variety’. (Zanettin, 2000:110).

With regards to data extraction, the concordancer is ‘the most common tool’

(Olohan, 2004:63). A concordancer works by finding all the occurrences of a word or term in a corpus and presenting the results in the form of concordance lines where the searched term is usually highlighted (example in figure 6.3 below) . ‘The length of the co-text displayed can be modified but is most conveniently viewed if it is limited to one line, or approximately 80 characters. This concordance display is also termed ‘keyword in context’ (KWIC)’ (ibid.).

121

Figure 6.3 Sample of concordances for unexpected from TEC extracted from Olohan, 2004:87

A specific feature of the research on proper names reported here is the fact that a new packet of tools is evaluated and ultimately employed to manipulate the corpus.

As will be described in the following sections of this chapter, BRIDGET is hosted by the University of Leeds and accessed online on the World Wide Web via an interface created by Serge Sharoff et al. at the Centre for Translation Studies (Sharoff, 2004).

Similarly, text processing tools, such as a multilingual concordancer, have also been developed and are also available to use with BRIDGET. Section 6.3 below assesses the steps involved in the creation of BRIDGET, the compilation and interrogation processes with a detailed description of the methods and tools used for the digitalization and aligning of the texts. Section 6.4 considers the software and processes used for interrogating the corpus, starting with an evaluation of the software used for the automatic identification and extraction of the proper names from the selected texts and following with a description of the multilingual concordancer used to find each name in the context in which they occur in STs and

TTs.

122

6.3 Creating BRIDGET

The opening sections of this chapter offered an overview of the issues and considerations around the compilation of a corpus. As this project sets out to describe the processes used to translate proper names from English Chick Lit STs into their corresponding TTs in Spanish and Italian, it is clear that a parallel corpus of Chick Lit novels is certainly the only data source that can be used. And such a specialized corpus has to be created anew. This corpus will be a valuable legacy of this project.

6.3.1 Design

With the research questions formulated (Section 4 of the Introduction) and the Chick

Lit fictional genre established as the source for the data (Chapter 5), the starting point is then the selection of the specific novels to be included in the corpus.

Consulting different Chick Lit websites7 and the catalogue of the British Library it was possible to compile a list of the British Chick Lit novels published in the UK since

1997 after the publication of Bridget Jones’s Diary8. The next step was to collect details on the translations available for these novels into the TTs of these study.

Information of all books published in a country can be found in databases held in their national libraries. These databases are nowadays digitalized and accessible

7 http://chicklitbooks.com/what-is-chick-lit/, http://www.chicklitclub.com/, http://www.chicklit.co.uk/articles/ 8 Helen Fielding’s Bridget Jones’ Diary, published in 1996, is considered to be the first novel in the genre, see Chapter 5. 123 online. For this project, information on STs was collected at The British Library while

Biblioteca Nacional Española and Biblioteca Nazionale Centrale di Firenze were consulted for information on TTs for Spain and Italy9.

Searching by the name of each Chick Lit author identified in the British Library online catalogue, an initial list with works from British Chick Lit authors was compiled. The

British Library catalogue allows the search by author, title or keyword, and the information displayed contains details on the book such as publisher, year of publication and ISBN. A similar search was then replicated in catalogues of Spanish and Italian libraries. The search engines work in much the same way and therefore, it was possible to recall bibliographic information in each language.

Books from the original list which were not found in the catalogues of both TLs countries were discarded. The final list amounted to 22 original titles translated into both languages under investigation. This constituted a total of 66 books, across a variety of authors, translators and publishers. (See appendix 1).

As Chick Lit is still a genre in its infancy all the novels in the list are controlled the copyright laws. Ultimately, it was the copyright holders who lastly shaped the list of books to be included in the corpus, as happens in so many CBTS projects (as seen in

6.2 with Saldanha (2011), Kenny (2001) or Bosseaux (2007), for instance). Initially, it was necessary to identify the copyright holders in each country for the books in

9 British Library: www.bl.uk Biblioteca Nacional Espanola: www.bne.es Biblioteca Nazionale Centrale di Firenze: http://www.bncf.firenze.sbn.it/

124 this list; again the catalogues of the different national libraries of each TL country were used. The Spanish government, through the Ministerio de Educacion y Ciencia, host their own on-line database of books edited in Spain. The search engine works in a much similar way to that of the libraries, but the results are presented by editions, therefore making it feasible to identify the original holder10 (example in figure 6.4 below).

Figure.6.4: Example of results from a search of ‘Bridget Jones’ in the Spanish database of books edited in Spain.

Still, issues around the selling of the translation rights, with the different imprints of the publishing companies and of different rules for different editions, meant that in some cases it was not even possible to ultimately identify the correct current owner of the rights.

As a starting point, letters in all three languages were drafted and sent to the publishing company which produced the first editions of each book. The Italian

10 http://www.mcu.es/webISBN/tituloSimpleFilter.do?cache=init&layout=busquedaisbn&language=es

125 publishers were the most receptive to the request, generally sending electronic files of the novels or, when not available, granting permission to manipulate the materials as needed. Publishers in the UK and Spain needed more persuasion.

Permission to scan (or in the case of Italian, to use the electronic copies) was finally obtained in all three languages for three novels: Jenny Colgan’s Amanda’s Wedding,

India Knight’s My Life on a Plate and Helen Fielding’s Bridget Jones’s Diary, which are the novels that compose BRIDGET (plots and bibliographical details included in

5.4).

These novels were published across a small time span (8 years), but crucially, not at the same time so to give an amount of diachronic variety. Similarly, they cover a selection of different authors, translators and publishing companies, so reducing the standardizing effect of a single author/editor’s preference. In addition, the different lengths and settings also offer an important degree of variation; for instance,

Amanda’s Wedding has a marked Scottish flavour as the main characters are originally from Edinburgh, therefore extending the cultural backdrop to the Scottish culture. My Life on a Plate is set within the more wealthy classes of society, and the connotations evoked through the proper names are very different to those found on

Bridget Jone’s Diary, for instance, which is set in a more working class environment.

All these factors contribute to a certain degree of variance across the texts while still holding enough comparability. Bibliographical information and a summary of the plots of the selected novels is included in Chapter 5 (section 5.4), where there is also a description of the special features of the Chick Lit genre.

126

6.3.2 Compilation

As described above, the publishing companies granted permission to use the texts but in most cases they did not provide electronic versions of the novels, therefore manual scanning and proof-reading was required. As already mentioned, the only exception to this were the Italian publishing companies (Mondadori and Salani

Editori) who were most helpful and obliged offering the electronic versions without any concerns.

For the digitalization of the rest of the texts a Cannon Lide 25 scanner was used.

The Omnipage Professional Optical Character Recognition programme (version 4.0) was used for the conversion of the images into texts. Omnipage offers valuable time-saving tools; for example, it would automatically identify the language of the text without having to change it manually every time a new language in being processed. It also has a user-friendly interface, that makes it easy to manage, offering at a glance view of all the scanned pictures as well as the data that has been converted into texts. The in-built spell-checking facilities are also useful, making it possible to proof-read and alter the texts on the spot, before the saving stage.

All these features contributed to a relatively smooth process. However, other factors such as the size of the pages, the quality of the paper or the line spacing sometimes made it difficult for the scanner to recognise the words and characters, requiring further manual checks and corrections and delaying the process.

127

The sequence established for the scanning of the novels was to start with the original book, followed by the Spanish translation and finishing with the Italian translation. Although there is no particular reason behind this order, the same process was maintained through the entire compilation and interrogation stages. The texts were saved in individual documents coinciding with each chapter of the book.

This was done in order to facilitate access to the data quickly (identifying one particular section or instance is done faster within shorter documents) as well as to minimize the damage in case of software crashes or hardware failures.

In order to ensure that the data can interact with any program or software, all the texts were saved as plain text with UTF-8 encoding throughout. In the first instance, they were kept in this format as a simple text document in Microsoft Notepad. This measure ensured that the format was consistent in all languages and texts, ready to be manipulated at the later stages.

As already mentioned in Section 6.2, a corpus which holds the same texts in more than one language is called a parallel corpus (McEnery and Wilson, 1996). A parallel corpus is, by definition, aligned. That is, ‘a segment of text in the TT can he identified as the translation of a segment of text in the ST or vice versa’ (Olohan,

2004:55); in other words, the different language versions have been made to match up at some level, typically the sentence or the paragraph, which one will depend on the type of research and the output sought. For this project the corpus was aligned at the paragraph level. The reasons for this are twofold, firstly it is possible to align

128 the texts at the paragraph level semi-automatically, using the find and replace features in Microsoft Word. For example, India Knight’s novel La mia vita as received from the Italian publishers contained the paragraph mark at the end of each sentence. Through a simple sequence of character replacements as shown in Table

6.2 below, it was possible to clean up the surplus marks very quickly.

1. replace all occurrences of “.^p” by “#”

2. replace all occurrences of “”^p” by “~”

3. replace all occurrences of “^p” by “ “

4. replace all occurrences of “#” by “.^p”

5. replace all occurrences of “~” by “”^p”

Table 6.2: semi-automatic alignment: replacement

Secondly, and more importantly, aligning at the paragraph level reveals more context when searching for a word in a concordancer. As has already been explained

(Introduction and Chapter 5), for this research the context surrounding each proper name is crucial for the understanding of the category of the name as well as the function that each name is performing in the text; aligning at the paragraph level readily presents this context for each name. As seen in Section 6.2, concordance lines generated with a concordancer generally offer a co-text of around 80 characters (Olohan, 2004:63) which was deemed not enough for the ultimate aims of this research. Although the majority of the programs available for corpus processing include an option to expand the results and add as much context as

129 needed, it was decided that having the context readily available when searching in the concordancer for each individual entity would work better for this particular project. Indeed, having the whole paragraph ready did prove to be very valuable at the data acquisition (section 6.4 below) and classification stages (Chapters 7 and 8).

All in all, the aligning of the paragraphs was the longest and most time-consuming part of the corpus compilation process. The method ultimately devised to perform the aligning was to use tables in Microsoft Word. The data originally saved as plain text UTF-8 in Notepad was manually copied and transferred across to a table created in Word, still maintaining the UTF-8 encoding. The table had three columns, the

English text was added in the first column to the left, the Spanish version to the second column in the middle and, finally, the Italian version took the last column to the right. Initially the paragraph mark in each text was used to create a number of rows according to the number of paragraphs in each text. Manual checks of each individual row and cell were performed to ensure that the same information was contained in each paragraph (the paragraphs do not necessarily automatically match in each language). When the information differed from language to language, the text was manually cut and pasted from one cell or row to another. This was repeatedly done until all the paragraphs were consistent in all languages.

In keeping with the process followed during the scanning, the texts were saved individually (one per each chapter of the book) as Word documents. All the files were then combined into one final document, saved in Microsoft Excel (using the

“select all/copy and paste” features in Microsoft Office suite). Excel is more useful

130 for dealing with large amounts of data; it also allows a better manipulation of the data, and a clearer production of figures and statistics, and has a simple interface which allows you to see the information in columns and cells and link different tables and sheets. The format used in this document (consistent with the format used in the individual Word documents described above) has the first column containing the

English text, the middle column the Spanish text and the third column the Italian text but instead of only one chapter this document contains the entire novel. A final check was performed to ensure that all the paragraphs were consistent and correct.

This Excel document containing the three versions of each entire novel with all the paragraphs aligned in the three languages was then sent to the server. As already introduced in section 6.2, the University of Leeds is the official host of BRIDGET, available as part of the corpus work led by Serge Sharoff (2004) at the Centre for

Translation Studies (accessed via www.corpus.leeds.ac.uk). It is important to mention, however, that due to copyright restrictions, BRIDGET is not readily available for the use of the general public. The publishing companies granted permission to manipulate and store the data in electronic format solely for the use of this research and similar ones undertaken by the Centre for Translation Studies or myself but were very clear that the texts were not to be made widely available.

6.3.3 BRIDGET

In summary, the result of the work reported thus far is a unidirectional trilingual, parallel corpus of original Chick Lit novels written in English and their corresponding

131 translations into Spanish and Italian. This corpus has been called BRIDGET in honour of the most popular character of the genre and contains a total of 652,926 running words (STs and TTs). BRIDGET is aligned at the paragraph level and the texts were originally left in raw form. The University of Leeds hosts this corpus, which can be accessed online through a purpose-built interface developed at the Centre for

Translation Studies of the same university.

6.4 Interrogating BRIDGET

As the main reason for using corpus in this project is the automatization of the retrieval process, it was decided that there was no need to parse, tag or mark the texts in any other form and consequently BRIDGET was left as raw text. After the aligning was completed, the corpus interrogation stage involved mainly work around the mechanical pinpointing, selection and extraction of the names from the texts.

The intention was to compile a list with all the names in the STs and then to find their translations in the corresponding TTs. The work with STs involved experiments with different retrieval tools (such as GATE: named-entity recognition, word- frequency lists) until a satisfactory 99% recall was obtained combining the outputs of these systems. The work with TTs involved mainly the use of a concordancer.

These tools and the methodology used to interact with them are explained in detailed in the following sections.

132

6.4.1 Identifying Proper names in STs

Taking into account the definition of proper name (Chapter 2, section 2.3) adopted in this study11 and the high population of elements which matched this definition contained in these books, the automatic identification of the items under evaluation from the texts was not as simple as originally anticipated.

A number of different methods for the automatic retrieval of proper names were tested. This evaluation was conducted via several experiments where the outputs of different systems were compared using measurements designed specifically to evaluate the performance of retrieval. Kenny (2001:102) stresses that ‘ultimately, the only way to evaluate the performance of (semi-)automatic means of retrieving instances (…) may be to compare the software’s output to a key arrived at manually by trained human experts’ (ibid.); following this idea, the performance of retrieval of these systems with BRIDGET was tested by means of a comparison of their outputs to a manually produced list of all the names in the first novel in the corpus,

Amanda’s Wedding. A list with all the names in this particular novel had been compiled as part of a previous research project (Cantora, 2006). The results obtained from these systems are compared to the manually produced list using measures such as recall and precision as well as the total numbers of true positives, false positives and false negatives identified by the systems.

11 Proper Name: a word, or group of words, that identifies an individual entity setting it apart from all the other entities in the same class and which, in the context and situation in which it is being used, has a specific referent and sense. Proper names both denote this referent and connote the characteristics which form its sense. 133

Fawcett (2006) defines these measurements as follows:

True positives (TP) are the proportion of answers to the query that are correct

False positives (FP) are the proportion of the answers to the query that are identified incorrectly

False negatives (FN) are the correct answers not identified by the system (that is, the names not identified)

Precision is the proportion of predicted positives which are actual positives: TP /

(TP+FP)

Recall is the proportion of actual positives which are predicted positive TP / (TP+FN)

In this particular project, the aim is to obtain as high a recall as possible, that is to identify all of the proper names in the text (or as many as possible). The amount of

‘noise’ found in the output, that is the words incorrectly identified, bear less weight.

In the first instance, I used GATE, a name-entity (NE) recognition programme developed at Sheffield University. This programme had successfully been used to identify NEs in journalistic texts with up to 96% accuracy in the results obtained.

The experiments with this tool, which are described in section 6.4.1.1., outlined, however, that GATE’s recall with modern fictional texts is low compared to the total number of names in the manually produced list. As a result, alternative methods for identifying the names in ST had to be devised. In the second instance, I attempted to use tags and word-frequency lists to identify all the words marked as PN (proper name). The recall of this method (section 6.4.1.2) turned out to be even lower than

134 that of GATE. IT help was thereafter engaged to develop a simple program to work with BRIDGET in identifying all the names in STs. This programme is described in section 6.4.1.3.

Combining the outputs of these three methods, I was then able to produce a list containing up to 99% of all the names in each SL book in BRIDGET. It is important to highlight that this list contains all distinct proper names used in STs. The list of proper names as obtained from each of these tools contained all the proper names found in the text, that is tokens of names (in other words, mainly one entry each time a name is used –e.g. the name of the main character appears a high volume of times-). It is to be expected that repeated occurrences of the same character’s name or the same place used over and over in one novel would normally receive the same treatment in the translation thus not offering new information for the analysis. The original list needs, therefore, to be pruned in order to retain only types: that is distinct proper names. However, this research works both with types and tokens. It is true that names with the same referent can be expected to be translated consistently but special care needs to be put in order to recognize cases in which a repeated name might be referring to a different entity. For instance, in the novel My

Life on a Plate, there is one mention of Elvis (Prestley), but there is also a dog called

Elvis, therefore the entry Elvis appears twice in the final list of distinct names from this book, with two very different referents: the singer and the dog. From that point of view, this project works both with tokens and types. When it came to identifying whether a name is pointing to a new referent alignment at the paragraph level as well as the multilingual concordancer (section 6.4.2.1) proved to be instrumental.

135

6.4.1.1 GATE

The Name-Entity (NE) recognition programme GATE [General Architecture for Text

Engineering] was developed at Sheffield University. GATE’s systems attempt to identify all instances of names of people, places, organisations, dates, job titles and monetary amounts. It is possible to run a text of 80,000 words through GATE in minutes. The creators of this tool claim that ‘NE recognition can be performed at

96% accuracy’ (Cunningham, 1999, Cunningham et al., 2002). In this context accuracy means the total of items correctly identified as a proportion of all of the items assessed, or in the terms of the definitions given above (TP + TN)/(TP + TN +

FP +FN). Where ‘TN’ is true negative, the incorrect answers not identified by the system (or the words that are not names that the system does not identify as names). In this project the number of TNs will be very high – as the vast majority of words in the corpus are not PNs and are not identified as such by any system.

Therefore the variations I am interested in between the numbers of TP, FP and FN will only create very small differences in accuracy; on those grounds, accuracy was not considered to be a significant measurement to use in this particular project. As explained above, as high as possible recall, that is correctly identifying as many names as possible, was the ultimate aim sought here.

Following the method explained, the results obtained from GATE are compared to the manually produced list using recall and precision. As already described above, precision is the proportion of the items retrieved that are correct answers to the

136 query. Recall is the proportion of correct answers to the query that are successfully retrieved.

The first step involved running GATE on the English text. The results are presented annotated in an XML document. A purpose-built XSL script is needed to show all the

NEs recognized by GATE in a list that can be viewed, and manipulated, on MS Excel.

With the help of an IT expert, I compiled this script and was able to generate this list in Excel (excerpt in figure 6.5 below).

Figure 6.5: Excerpt from the purpose built script for GATE.

GATE found a total of 2,585 NEs, i.e. tokens. Through a manual clean, a total of 299 types were quickly confirmed. The manually compiled list comprised 455 types.

137

Of those 299 types identified by GATE, 215 were true positives: present in both the manually generated list as well as in the GATE list. On the other hand, 84 were false positives: NEs wrongly identified by GATE as proper names. Of these Chuck Alex or

Aren’t Posh are two examples. Finally, GATE failed to identify 240 proper names that form part of the manual list and are relevant for the analysis. This means that on

Amanda’s Wedding GATE achieved 71.66% precision and 47.25% recall.

The biggest problem comes with the low recall, which means that there are a lot of proper names in the text that are not identified as such by GATE. Considering the specific categories that GATE is looking for as NEs in a text, (that is people, places, organisations, dates, job titles and monetary amounts), only people, places and organisations apply directly to the categories of proper names found on BRIDGET

(Chapter 7). Out of these three categories, GATE identifies most of the entities.

Consequently, the problem arises with the other categories (product names, book or film titles, festivities or celebrations, etc.), which, of course, GATE was not geared to look for. It was clear, therefore, that alternative automatic retrieval processes needed to be considered.

6.4.1.2 Word-frequency lists

Word-frequency lists present the whole lexical range (each different word) that can be found in a particular text. The results are generally ranked by the frequency with which the word is used in the corpus and are also often marked with useful information such as, for instance, the lexical category of each word (Olohan,

138

2004:77). With manual checks of these lists, it should then be feasible to identify which words are proper names.

The effectiveness of this method was tested, following the same pattern used with

GATE’s retrieval, through a comparison of the manually produced list (Cantora,

2006) of all the names in Amanda’s Wedding against the output of the word- frequency list for this novel generated using the corpus tools available at the Centre for Translation Studies in Leeds.

The tool presents the results online in HTML format in four columns as represented in figure 6.6.

Figure 6.6: Excerpt from frequency list for Amanda’s Wedding

139

The first column shows the rank of the word in the text (starting with the most frequent). The second column represents the number of times the word is used in the text. Column 3 lists the words themselves. Finally, the fourth column shows the category of the word (preposition, verb, noun, etc.). Total numbers of types and tokens are also included.

Several steps were involved in processing the data: firstly the HTML online list was imported into WordPad with a view to converting the data into plain text. This data was then transferred into an Excel spread sheet creating an Excel document replicating the word-frequency list found online. This list is then revised looking for all the words marked as PN (proper name) as well as those words which appear capitalized (even if they have not been marked as PN) as they could be part of a compound name (a book title, for instance). Working is this way it is possible to create an initial list with candidate proper names in the corpus. With the aid of the concordancer (searching for the words in this list of potential proper names in the corpus) a final list of names occurring in the corpus is created.

The original frequency list comprised 2,869 tokens, that is occurrences of each word marked as PN and out of these 252 types, that is distinct proper names. Out of these 127 were true positives and 125 false positives. The manually produced list of names contained 455 distinct proper names, therefore leaving a total of 330 proper names not found. This represents 27.47% recall, and 49.60% precision.

140

The main problem identified were the compound names formed of two or more words (Tom Cruise, for example) which appear separated in the list. It was necessary to identify the elements that form the compound and unite them as one item in the final list. This can be done with the aid of the concordancer, but it is a delicate and time-consuming task. Another problem that arises as a consequence of this truncation of the compound names is with the total numbers of types and tokens which is distorted by the process. Take the case of London, this entity appears a total number of 28 times. This includes the cases when London is used as part of a compound name, such as London Zoo or East London referring to a different referent. So the count of 28 times is valid for the number of times the word

London is used but not for other occurrences of this proper name with a different referent. There are numerous examples of this problem in the list of names from

Amanda’s Wedding:

East: East London, East Coast, East End

Little: Little Chef Burger, Little Dave

Man: Man U., Action Man

St: Rue St. Denis, St. Germain, Monsieur St. Laurent

Father: Father Bernard, Father Christmas, The Sash My Father Wore

Mary: Queen Mary’s, Holy Mary, Bloody Mary

Love: Toddlers Love Learning, Who Will Love My Children?

Dolce: Dolce Vita, Dolce & Gabanna

Hill: Heywood Hill, Notting Hill Café

Billy: Billy Goat Gruff, Billy No-Mates

141

In view of the low precision and the even lower recall of this method and especially taking into account that those figures were obtained after an arduous process of manual clean and guess work using the concordancer, it was evident that this method of retrieval was not effective.

Alternative methods were considered, for example, identifying all those words whose

(normalised) frequency in the corpus is higher by some threshold than their frequency in e.g. the British National Corpus. That will pull out all the NEs and some other words, but it should be easy to filter the list. However, there was again a possibility that this retrieval process would not bring out the desired results. The decision was taken then, to enlist some IT assistance in designing the specifications needed to build a simple program specifically for identifying proper names in literary texts12.

6.4.1.3 Access for BRIDGET

Access for ‘BRIDGET’ is the name given to this program specifically developed for automatically identifying and extracting proper names from the corpus (appendix 5).

The algorithm used is based on the typical characteristic of proper names of being written with an initial capital letter. As seen in Chapter 2, this is one of the

12 I would like to acknowledge the assistance of my partner and colleague Mr. Grant Cocks, (Cocks, 1993, Cocks, 2004) in teaching me Java CC. Grant oversaw the development process of ‘Access for Bridget’ and applied his deep knowledge of data architecture to the design of the application. 142 referential criteria generally used in the linguistic definition of proper names. This characteristic has been ruled out as a universal descriptor of proper names (Algeo

(1973), Jespersen (1924/1964), Fernandez Leborans (1999)), as it does not apply to all languages; for example, in German both common nouns as well as proper nouns are given an initial capital letter. This characteristic does not form part either of the semantic definition of proper names adopted in this study; however, it still complies to the grammatical rules of all three languages of this study: English, Spanish and

Italian, where proper names almost always start with an initial capital letter and therefore could work as the pattern for the program to look for. There are, of course, exceptions to this rule such as eBay, for instance, but these examples are very rare and in general it can be expected that authors would follow the grammatical rules and conventions of the languages around proper names.

The code was written in JavaCC and it consists of a parser which assigns a structural description to raw text, identifying sentence boundaries, direct speech and other quoted units and a set of commands which carry out different actions for words occurring at different points within those structures. The program finds all the words or sequences of words in the text with an initial capital letter, with the exception of single capitalized words occurring at the beginning of a sentence (as identified by the parser), and stores those words in a relational database (Microsoft Access). The raw texts, aligned paragraph by paragraph in all three languages, are held in one table. Another table holds the words identified as proper names and a third table keeps the relations between those lists of names and the paragraphs where they occur in the original table holding the texts. The user can query the database with a

143 name and retrieve all the paragraphs in which that name occurs in the data table. It works much as a normal concordancer would work, bringing up all the paragraphs in which the searched proper name occurs together with its aligned corresponding matching paragraphs in the other languages. However, unlike a concordancer,

Access for BRIDGET only allows the user to search for words included, and as included, in the list of proper names identified by the parser. A search for other words or collocations is not possible. For example, Elizabeth Arden Eight-hour Cream has been identified as a string of capitalized words forming one compound proper name (Chapter 2). It is possible to search for this name and find all the paragraphs where Elizabeth Arden Eight-hour Cream occurs, but the results will not include occurrences of Elizabeth Arden. If Elizabeth Arden occurs as a distinct proper name in the text and can be found on the proper names table, it would then be possible to perform another search for that name. If it does not occur as a distinct proper name, then the search does not bring up any results in the proper name query window. This could potentially be a problem when trying to identify whether one particular word is a proper name or not, but it is not a problem when searching for particular instances of proper names identified as such.

Teething problems quickly emerged after generating the first results using Access for

‘BRIDGET’:

- The program identifies words or strings of words which are capitalized within each sentence in the texts. In order to avoid the first word of a sentence, after a full stop, to be selected as a proper name in every case, the program ignores a capitalized

144 word occurring at the beginning of a sentence. Effectively, this means that a one word proper name which occurs in the text only once at the beginning of a sentence would be missed out by the programme. This particular scenario is not likely to occur repeatedly. The loss incurred by missing out the odd name which appears in the text only once is far less than the advantage of eliminating ‘noise’ with a proper name (I refer back to the work undertaken with Word-frequency-lists described in section

6.4.1.2 above).

- Compound names that include a non-capitalised word such as & or and (e.g.

Pitcher & Piano, Furrow and Ball) are not identified as such. In a similar way to what happened with Word-frequency lists and proper names formed of more than one word, it is necessary to manually alter the list combining those proper names. The number of occurrences of these compound names is considerably lower, though, than the total number of compound names that needed reconstruction in the previous method, and therefore it can be said that an improvement has been made, albeit the problem has not been completely eradicated.

- Proper names occurring as the second word of a sentence, for example But Robert,

And Charlie, are taken to be a unit, as a string of capitalized words similar to Tom

Cruise, and included them as distinct proper names.

A further process was devised in order to solve the last two problems through a manual check of the original list. Microsoft Excel commands were created to automatically recombine the elements of a proper name which had been split or to

145 rename any name that has been misinterpreted or misspelled, without corrupting the total numbers of occurrences for each name.

Opening in Excel the list of proper names originally identified by the programme and their unique reference number assigned by the programme, three more columns are added, Rename with, Combine with, and False Positive (figure 6.7).

Figure 6.7: Example of the table used to modify results from ‘Access for BRIDGET’

Through a manual check of the list, those names representing those problems are marked in the corresponding column. If the name needs to be renamed the new name is written in the column Rename with. If the name needs to be added to the count of another name, as is the case of And Charlie for example, then Charlie’s unique reference number is added to the column marked as Combine with. This

146 name will be renamed as the name corresponding with the unique reference number introduced in the cell and added to the count of that name. If the word identified by the programme is not part of a genuine proper name occurring in the book then it is classed as a false positive. It is then possible to run some commands on the original list of names in Access which will alter this list with the new names/numbers.

In keeping with the experiments performed with GATE and Word-frequency lists, the effectiveness of retrieval of Access for ‘BRIDGET’ was tested through a comparison of the output for Amanda’s Wedding against the manually produced list of names in this novel.

The total number of tokens identified by the program as proper names was up to

4899 out of which 683 were distinct. Within the distinct proper names 102 were false positives, 67 had to be renamed with, 146 had to be combined with, and 435 were identified correctly and can be classed as true positives. 20 proper names were missed out. This means that Access for ‘BRIDGET’ had 95.60% recall and 63.68% precision. This results show an improvement from the previous methods, although manual checks were still needed. However, the process for introducing the changes

(that is the combine with and rename with) is clear and easy to follow and can be done at a relative speed. More importantly, the almost 96% recall achieved, makes

Access for ‘BRIDGET’ the best tool to use with this corpus.

Table 6.3 below collects all the measurements of each retrieval method tested:

147

Amanda's Wedding Access for GATE Word List BRIDGET

Tokens 2585 2869 4899

Types 300 252 683

False Positives 84 127 102

Rename With n/a n/a 67

Combine With n/a n/a 146

True Positives 215 125 435

Total PNs in text 455 455 455

Not identified 240 330 20

Recall 47.25% 27.47% 95.60%

Precision 71.66% 49.60% 63.68%

Table 6.3. Outputs of the different retrieval methods used with Amanda’s Wedding

6.4.2 Identifying Proper Names in TTs

6.4.2.1 Concordancing

As outlined above (section 6.2), a concordancer identifies all the occurrences of a word, or group of words, in a text. The software usually has an interface with a search box where you input the item that you want to find and the program brings up all the instances in which the said item appears in the corpus. With aligned parallel corpora both STs and TTs come up in the results. The idea with BRIDGET was to use a concordancer to search for each individual name identified in the STs

148 and, as the text is aligned at the paragraph level, to then look for the translation used with each name in the TTs.

The multilingual concordancer developed by Serge Sharoff (Sharoff, 2004) as part of the corpus tools created at the Centre for Translation Studies was used. As well as for storing and easy access purposes, this multilingual concordancer offers the extra bonus that it can interact with several languages at once, a feature that particularly suits the trilingual character of BRIDGET. Most of the concordancers available on the market allow search in two languages, but a special feature of the concordancer developed at Leeds is that it allows the user to search for a word or selection of words in the ST, or in any of the TTs, any combination is possible.

Figure 6.4: Sample concordances for ‘amanda’ from the multilingual concordancer

As BRIDGET is aligned at the paragraph level, the results are presented in three columns, showing all the paragraphs in which the word or words appear in the selected search language as well as their corresponding matching aligned

149 paragraphs in the other languages. Identifying the translations used with each name is then straightforward.

This multilingual concordancer proved to be an extremely useful tool, not only as a way to finding the translations of the SL names, but also for the classification of the data and the procedures (Chapters 7 and 8). It certainly is a time-saving tool, which allows one to find any instance of any name in seconds, a task which would be near- impossible to perform manually.

6.5 Conclusions

In this chapter I have described the process for the design, compilation and interrogation of an electronic corpus, and linked the concepts directly to the processes used in the creation and interrogation of BRIDGET.

Although the process of digitalizing the texts and aligning the corpus is long and time-consuming, the benefits from having the texts in electronic format as well as those obtained from the use of software tools from corpus linguistics outweigh the negatives. These corpus-processing tools facilitate the manipulation of large amount of data which otherwise would be impossible to process manually. Furthermore, the availability of the aligned texts in the concordancer allows for the searching of the names at great speed as many times as it is required. The concordancer was instrumental also at the analysis stage, when it was often necessary to go back to

150 the context of each name and to recall the information several times, for instance for identifying the translation procedures employed with each name (Chapters 8 & 9).

With all the data acquired, and the methodology explained, Chapters 7 and 8 move on to cover the analysis stage. Chapter 7 presents a taxonomy for proper names, as a way to offer order and consistency to the data. Chapter 8 covers the map of translation procedures used to classify the translation instances. Each proper name identified through the corpus tools and the methodology described in this chapter was classified according to the taxonomy created (Chapter 7) and a translation procedure was assigned to each translation instance (Chapter 8). The combination of these culminates in Chapter 9, where the results of these processes are combined and the patters for the translation of proper names in Chick Lit uncovered.

151

Chapter 7. Data Analysis 1: Classifying the Proper

Names.

“They call me ‘Wonderful’ so I am wonderful, in fact… it’s so much who I am it’s part of my name”. Schwarts, S.: Wonderful – Wicked the Musical

7.1 Introduction

The results of all the data identification and extraction processes described in

Chapter 6 come together in three Excel documents (one per book in the corpus) each one containing three columns, one per language, with all the distinct names found in STs and their corresponding translations in TTs. In keeping with the format used throughout the data acquisition process, the order follows the SL in the first column to the left, the Spanish translations in the second column in the middle, and the Italian translations in the third column to the right. A full copy of these documents can be consulted in Appendixes 2, 3 and 4 of the enclosed CD. These lists of names will be the base of the analysis process, as described in the following sections of this chapter as well as in Chapters 8 and 9. A total of 1,610 distinct proper names were identified in the three SL texts. The first stage of the analysis involves the classification of all the data. In the first instance, the names are presented in alphabetical order, from the SL point of view. However, a more comprehensive classification was devised in order to offer structure and consistency to the data. Furthermore, the categories envisaged in the taxonomy will be combined in Chapter 9 with the classification of the procedures (Chapter 8) for a more exhaustive analysis of the relationships between each type of name and the

152 procedures used to translate them. The combination of taxonomy and procedures to study translated proper names represent a novelty approach and is one of the major contributions to knowledge of this study. The methods devised for the analysis

(section 7.3) can be generalized and taken further to be replicated in other studies on translated proper names; furthermore, the comprehensive taxonomy is general enough to be used to classify names extracted from different text types as well as in projects from other disciplines.

The taxonomy presented here adopts a faceted form; Chapter 5 (section 5.2) described how Chick Lit novels are based in modern society and how they attempt to recreate a world familiar to the audience through the use of proper names from their real world; from this point of view, a classification of names which takes into account these particular point has a clear advantage. As well as structuring the data, the aim of this classification is to highlight the relationships between the names, as well as the relationship of the names with the real world outside the novel. To that aim the taxonomy takes a faceted form; that is, it is comprised of different facets, or layers, each one relating to one of those features. When combined, all the facets give an overview of the particular features involved with each individual name.

Traditionally, proper names have been classified by Onomastics in a very rigid way, similar to the rigid bibliographic classifications used during the early part of the

Twentieth Century. These traditional classifications aimed to unidirectionally categorise all possible types of names available, that is, people’s names, names of places, of things, of materials, and so on and so forth. The classification proposed

153 here goes a step further, taking inspiration from the more complex faceted classifications used in information science since the 1960s and 1970s. A faceted classification categorises different aspects (or facets) of a subject field; these different facets can be combined depending on the information sought, a feature which offers more flexibility and adaptability to the classification. The taxonomy proposed here is comprised of four facets: Ontological, Semantic, Scope and Sphere.

The Ontological facet categorises the different types of names in a similar way to the traditional classifications used by Onomastics. The Semantic facet, moves to consider the semantic characteristics of the names, whereas the Scope and Sphere facets work together to establish the relationship of the names with the world outside the novel and its reality. It is the combination of all these facets into one taxonomy which offers real flexibility and novelty. With a faceted classification it is possible to assign multiple sets of attributes to an object making it feasible to navigate the information along multiple paths. This makes this type of classification extremely useful for this particular study into the translation of proper names as it facilitates the combination of each attribute of the name with the procedures used to translate that particular type of name.

Each facet is described in detail below. The final section of this chapter covers the methodology and processes followed for applying this taxonomy to the list of proper names extracted from the corpus, firstly to ST names and then to TT names; the section also explores the end results and their usability and reliability of the method for this research and similar ones.

154

7.2 Taxonomy for Proper Names

7.2.1 Ontological Facet

7.2.1.1 Introduction – The ontological classification of the Prolex Group

As introduced above, the Ontological facet categorises the different types of names that can be found in BRIDGET. It basically represents an organized schema of subject categories. If a proper name is understood as a word which refers to one entity (Chapter 2, section 2.3), this facet collects the different types of entities that a proper name can refer to. In other words, when we think of a proper name we think of the name of a person, the name of a place, the title of a book, and so on; this facet attempts to categorise all these different possibilities. The primary aim of this facet is to structure the data. In addition, it offers a clear terminology to be linked to the analysis of the different procedures. The final aim is to explore the relationship, if any, between a particular type of name and a translation procedure; section 9.3 will combine the different types of names identified in this taxonomy, with the procedures employed to translate them. This will be done with an aim to identify any possible pattern between type of name and translation procedure more frequently used with it.

The categories in this facet have been devised primarily following the classification established by the group Prolex. This group, based at the University of Tours

(France), is working on the creation of an electronic multilingual dictionary of proper names. They work directly with real proper names from real texts, in an attempt to categorise proper names and their translations. The ultimate aim is to structure a

155 multilingual dictionary which can be used in Machine Translation. Originally, they envisaged a hierarchical relational database (based in the classification presented by

Bauer, 1985), taking into account the different relations between the names (Grass,

2002, Grass et al., 2002, Grass, 2000) with two levels: hypertypes at the top and types underneath. In 2004 (Grass et al., 2004, Grass et al., 2006) the group announced the adoption of a faceted structure, founded on a four level ontology: 1- level of instances (the proper names as such), 2- the linguistic level (prolexemes), 3- the conceptual level (numerical pivots) and 4- the meta-conceptual level (types and hypertypes). These levels represent in a relational database the different relations between the names. The facets adopted by Prolex very specifically target the aims of structuring and creating a multilingual dictionary. It is, nevertheless, significant that the group opted to modify the structure of their database into a faceted one. This can be taken as a further reinforcement of the benefits of applying a faceted classification as opposed to a more rigid directional one.

Although some of their facets very clearly apply directly to their own aims, for instance the level of instances (their own lists of names) or the conceptual level (the numerical values that they assign to the data in their database), the meta- conceptual level could potentially have further applications. From the standpoint of their work being based on real translated proper names, and this facet consisting in different categories for structuring proper names, their categories at the meta- conceptual level could potentially also be applied to the names in BRIDGET.

Therefore, as a starting point, these categories (described below) were used as a first approach to the data extracted from the first book in BRIDGET, Amanda’s

156

Wedding. Section 7.3 below describes more in detail all the process of applying this classification.

The meta-conceptual level of the Prolex ontology is divided into conceptual, homogeneous lexical fields, which help to organise the data contained in a structured relational database. The higher level is represented by ‘hypertypes’ whereas the different sub-categories to be found inside each hypertype are termed

‘types’. Table 7.4a below shows the hypertypes and types in the classification of the

Prolex group.

157

Hypertypes Types First names Celebrities

Dynasties

Anthroponyms Divinity, mythical or fictive Firms Political Parties Artistic Ensembles -including Sports and music-

Public or Private Institutions -Universities, Hospitals, Institutions- NGOs and international bodies Inhabitants Country Hierarchic Region Classification Town and villages

Group of countries Toponyms Quarter, Road, Street Buildings -including parks, gardens, theatres and bridges- Hydronyms Geonyms Celestial Objects Fictive or mythical places Private or public institutions Brand name or Trade Mark Firms Ergonyms Books films, theorems, etc. Mythical object names Vessels Decorations Meteorological event Historical or political event Prognonyms Sporting or cultural event Feast name with cyclic character Table 7.4a: Classification of Proper Names by Prolex Group, 2002.

158

From the classification of the names in Amanda’s Wedding according to the system of the Prolex group it ultimately emerged that different categories could be devised for a more in depth coverage of the different entities that can be referred to by a proper name. Furthermore, the aim of Prolex is to create an electronic dictionary of proper names and it quickly transpired that the categories proposed by the group could be more exhaustively tailored into a broader schema of subject categories, which would translate into a more comprehensive ontology. As described in chapter

5, the Chick Lit genre uses modern proper names in a very particular way; making reference to a wider amount of entities than those covered by Prolex’s ontology. It was the case that some of the entities in BRIDGET were not included in the Prolex taxonomy (details below). Therefore, it was pertinent to adapt some of the categories in the Prolex group taxonomy. The idea was that the greater the number of categories in the map the more comprehensive the taxonomy would be. This opens up the possibility of exploring the relationship of each individual and precise type of entity with the translation procedures employed to translate them (Chapter

9). The wider the categorization of the entities, thus, also the more comprehensive the analysis of the procedures employed with each category.

The end result of this data-driven approach is an ontological facet which includes categories which cover a wide range of different types of proper names that can be found on modern society, and modern literature, nowadays.

The Ontological facet ultimately used with all the names in BRIDGET is described in section 7.2.1.2 below.

159

7.2.1.2 BRIDGET’s ontological facet

Following the Prolex group, the ontological facet of the taxonomy adopted in this study has likewise been structured by hypertypes and types. Similarly to top-level categories of information science ontologies, the hypertypes represent the primary concepts behind the names, while the types specify the concrete field that the name belongs to. Five categories are included within the hypertypes:

1. Anthroponym primary concept living thing

2. Toponym primary concept place

3. Ergonym primary concept object

4. Pragnonym primary concept event

5. Chrematonym primary concept group

Each hypertype is described in detail below under individual headings. Each section starts with an introduction followed by the description of each type contained within.

This explanation is followed by a table containing a summary description of the types for each hypertype with an example from the corpus given for each. Some categories include a note to offer further explanations of their nature and their role in the taxonomy.

160

1. Anthroponyms – living thing

The hypertype Anthroponym covers the names representing the primary concept of living thing. This is mainly names of people but names of other alive creatures are also included in this hypertype, for instance names of animals or plants. The types for this hypertype are as follows:

First name: Personal name given to someone at birth and used before a family name.

Amanda, Bridget, Daniel, Fraser and all the other character names in the books.

Surname: hereditary name common to all members of a family.

Jones, Cleave, McConnald and all the surnames of all the characters in the books.

Full name: First names and surname. Covers all references to a person using both their given name plus their family names.

Fraser McConnald, Amanda Philips, Daniel Cleaver.

Entertainment figure: Names of people dedicated to the performance world or to entertain others.

It covers singers, actors and actresses, models, celebrities, etc.

Britney Spears, Kate Moss, George Cloony.

161

Cultural Figure: Names of intellectual people. It covers scientists, politicians, writers and novelists, authors, etc.

Stephen Hawkins, Margaret Tatcher.

Nickname: Familiar or humorous name given to a person instead of their given name. It also covers for pseudonyms.

The Gustard (name often used by the characters to refer to Fraser’s brother Angus).

Title: Describes someone’s position or job. Includes honorific posts.

Laird (title inherited by Fraser from his grandfather).

Right Hon (used by Bridget in certain occasions to refer to members of parliament).

Zoonym: names of animals.

Covers for any deictic reference to pets, zoo inhabitants or any kind of name given to an animal.

Elvis (the dog of Clara’s step-father).

Or for instance Niebla, Heidi’s dog, or Copito de nieve, the white gorilla at

Barcelona’s zoo (examples not from the corpus).

Botanic: Names of plants and trees.

Casablanca Lilies (in Clara’s step-father’s house).

Ethonym: Collective name or adjective derived from a proper name.

Rubenesque (used by Bridget while wondering about women’s figures)

162

Types Explanation Example from Notes BRIDGET Given name Amanda First name Family name Ms Philips These types Surname categorise the concrete field of Given name plus Amanda Philips personal name Full name family name McConnald

Applies to Tom Cruise The distinction Entertainment individuals in the between these two Figure entertainment types was deemed world pertinent as it helps Writers and other Margaret Thatcher separate, for cultural, historical example, Britney or political Spears from Stephen figures. Also Hawking or Jane Cultural Figure includes members Austen. Once again,

of the royalty, each category could dynasties, etc. potentially receive different treatment in the translations Familiar or The Gustard humorous name

Anthroponyms given to a person Nickname instead of their given name. Includes pseudonyms. Described Right Hon someone's Title position or job. Includes honorific posts Names of animals Elvis (the dog) These two types are Zoonym included in the Anthroponyms hypertype due to Names of plants Casablanca (lilies) Botanic their condition of being alive. Collective name Rubenesque or adjective Ethonym derived from a proper name. Table 7.5: Anthroponym Hypertype with its types

163

2. Toponyms – place

The hypertype toponym covers the names of places. This hypertype includes from

Geographical extensions to man-made buildings or locations. The types described below are ordered from the wider ranging to the more concise.

Celestial: Names related to the sky or outer space. Astrological. Covers for names of planets, galaxies, starts, etc.

Venus.

Continent: Names of any of the world’s main continuous expansions of land.

Europe.

Country: Nation with its own government occupying a particular territory.

Portugal (place where Bridget’s mother elopes to).

Region: an administrative district of a city or country. It includes names of bigger parts of a country such as the counties or regions in England or the comunidades autónomas in Spain.

Surrey (Region where Amanda comes from)

Madrid, Andalucia, Babaria (examples not from the corpus).

City: a large town

London (where the characters live).

164

Village or town: a built up area with a name, defined boundaries and local government and generally smaller than a city.

Woking (town in Surrey where Amanda comes from).

Locality: an area or neighbourhood

Stoke Newington (which Clara describes as a part of London with its own personality).

Transportation: Refers to names of entities related to travelling and covers names of airports, underground lines or train stations, names of highroads, etc.

The District Line (which Melanie takes to go to work).

Hydronym: Names of bodies of water, includes names of oceans, seas, rivers, lakes, etc.

The Thames (the river in London).

Institution: Names of schools, universities and other educational centres.

Durham University (where Amanda applied to).

Entertainment premise: Places where people go for socializing. Includes names of bars, restaurants, cinemas, clubs, etc.

The Ivy (posh restaurant in London where Clara often goes for dinner with her mother Kate).

165

Retail premise: Places where people go for purchasing goods or services. Includes names of hairdressers, shops, supermarkets, etc.

Tesco (where the characters often do their grocery shopping).

These two categories, entertainment and retail premises represent one of the main novelty of this categorization as the distinction between these is rarely included in proper name taxonomies. These entities are considered toponyms from the point of view of generally being used with verbs of movement. The distinction between these two is deemed necessary to be included in a comprehensive taxonomy for proper names as they represent different kinds of concepts and entities. From this standpoint it can be hypothesised that different translation procedures could be used to transfer them into different languages and cultures.

Monument or building: Names of erected structures

Xyler Building (from where Alex sent Melanie a post card).

Theoretical place: Intangible spaces such as Utopia. It also includes imaginary places.

The Land of Alex (where Fran thinks Alex lives, in his own world, without taking into account anyone else).

166

Types Explanation Example from Notes BRIDGET Celestial Names related to the Venus sky or outer space. Astrological Continent Any of the world's Europe main continuous expansions of land

Country Nations Portugal Region Bigger parts of a Surrey country such as the counties or regions in England or the comunidades autonomas in Spain City Large towns London Village or town Smaller towns Woking Locality Includes streets, Stoke Newington roads, neighbourhoods, boroughs, etc. Transportation Includes entities The District Line related to travelling, for example names of airports, underground

Toponyms or train stations Hydronym Names of bodies of The Thames water Institution This category covers Durham University for names of schools, universities or other educational centers

Entertainment Places where people All Bar One premise go for socialising (bars, restaurants, cinemas, clubs, etc.) These two categories are considered as toponyms, due to their nature of being Retail premise Places where people Tesco locations generally used go for purchasing with verbs of movement. goods or services (hairdressers, shops, supermarkets, etc.) Monument or Erected structures Xyler Building building Theoretical place Intangible spaces such The Land of Alex as Utopia, or imaginary ones. Table 7.6: Toponym hypertype with its types

167

3. Ergonyms – Object

The hypertype Ergonym covers for names of objects and man-made products. This includes brands and names of manufactured goods that the characters consume. It also includes titles of the books they read, the songs they listen to, and so on. The types corresponding to this hypertype are as follows:

Brand/product: Names of manufactured goods. Includes the names of all the

products and things that people consume.

Ryvitas.

Book/Theater /Poem: Names of literary works, including titles of novels, plays, or

specific poems.

Pride and Prejudice (mentioned several times in Bridget Jones’s Diary).

Film/tv: Names of entertainment programmes or films.

EastEnders (often watched by all the characters in all the books).

Song: titles of songs and music albums.

Do they know is Christmas? (song that the characters listen to at Melanie’s dinner

party).

Newpaper/magazine: Names of publications which contain at least one of the

following sections: news, articles, advertisements and correspondence.

168

The Times (Newspaper that Clara’s husband reads) or Panache (magazine where

Clara works as a journalist).

Work of art: Names of artistic works

The Guernica

Game: Names of means of passing time

Spin the bottle (game played during Melanie’s dinner party).

Academic: Names of subjects, degrees, academic awards, etc.

A level French (achieved by Clara’s step-sister).

Individually named object: Names given to individual objects. Includes vessels or names given to cars or a lorry as well as names given to a child’s teddy bear.

Bunny (the name of the soft cuddly bunny used for comfort by Clara’s son).

169

Types Explanation Example from Notes BRIDGET Brand / Product Names of manufactured Ryvitas goods Book Names of literary Pride and Theatre Works Prejudice Poem Film/TV Names of entertainment EastEnders

programmes

Song Do they Know Song Titles it's Christmas? Newspaper Names of The Times

Magazine publications Ergonyms Work of Art Names of artistic The Guernica Works Game Names of means of Spin the Bottle passing time Academic Names of subjects, degrees, academic A Level French awards, etc. Individually Names given to named object individual objects Bunny (for instance, a child's teddy bear) Table 7.7: Ergonym hypertype with its types

4. Pragnonym – Event

The hypertype Pragnonym covers for the primary group event, this is names of different events such as festivities or of historical or sporting happenings. The following types are included:

Historial / political event: Significant events that occurred in history

Holocaust (Bridget would like to know more about this subject).

170

Sporting / cultural event: Significant sporting competition or cultural occasion.

The Queen’s Christmas Message.

Cultural and sporting events are included in the same category. There is no reference to sporting events in BRIDGET but it is important to include it within the pragnomyns (events) as it could appear in other books or text types. This ensures the usability of the taxonomy in other studies.

Festivity: Names of celebrations, parties or fiestas

Una and Geoffrey Alconbury’s New Year’s Day Turkey Curry Buffet (party at the house of Bridget’s parent’s friends which she attends every New Year).

Time Period: Specific time of the year or period

Lent.

Illness: diseases or period of sickness

The Colditz Fever (reference made by Amanda).

171

Types Explanation Example from Notes BRIDGET Significant events Historical/political that occurred in Holocaust

event history Significant sporting The Queen's Sporting/cultural competition, or Christmas event cultural occasion Message

Pragnonyms Festivity Celebration Christmas

Specific time of the Time Period Lent year, or period Diseases or period Illness Colditz Fever of sickness Table 7.8: Pragnonyms hypertype with its types

5. Chrematonym – group

The hypertype Chrematonym covers the primary concept group. It encompasses associations, groupings or gatherings of people. This hyptertype was not included by the group Prolex (they included some associations into the anthroponym and ergonym hypertypes). The concept of a hypertype to cover specifically for groups was taken from Bauer’s (1985) ontology (originally used by Prolex group). Although

Prolex decided to cancel this hypertype in their classification, the decision of restoring this category was taken on the basis of it covering a wide range of entities, different from the specific individuals covered under the anthroponyms. As already mentioned, this will contribute to completing a comprehensive taxonomy which would cover as wide a variety of entities as possible. The different types included in this hypertype are summarised below:

172

Firm: names of business and commercial companies

BP (company that Angus works for).

Non-Governmental Organisation / charity: Organisations set to provide help.

The Samaritans (Melanie often considers phoning them).

Arstistic ensemble: Bands, music groups, orchestras, and any association related to music.

Queen (band that one of Melanie’s lovers is quite fond of).

Sport club: Names of sporting teams, groups, etc.

Man U. (football team where the favourite player of Clara’s son plays).

Public or private institution: Organisation founded for a religious, educational, professional or social purpose

The Met (One of Amanda’s friends has a cousin at The Met).

Political party or organisation: Relating to the government or public affairs

Labour (Bridget talks about members of government).

Art/literary movement: Literary or artistic periods, shares of ideas, etc.

Nouvelle Vogue (mentioned by Clara’s step-sisters).

173

Religion: System of faith or worship

Orthodox Jew (mentioned by Bridget while thinking about the Holocaust).

Informal group or association: A group of people organised for a joint purpose

Moonie cult (referred to during Melanie’s dinner party).

Example from Types Explanation Notes BRIDGET Firm Business BP

Organisations Non-Governmental set to provide Samaritans Organisation/charity help Bands, music Artistic ensemble groups, Queen

orchestras, etc. Names of Sport club sporting teams, Man U.

groups, etc.

Organisation founded for a Public or private religious, The Met institution educational, professional or social purpose Chrematonyms Relating to the Political party or Coldstream government or organisation Guards to public affairs Literary or artistic periods, Nouvelle Art/literary movement shares of ideas, Vague etc. System of faith Religion Orthodox Jew or worship A group of Informal group or people organised Moonie cult association for a joint purpose Table 7.9: Chrematonym hypertype with its types.

174

These descriptions and tables collect all the types included within each hypertype in the ontological facet. The primary purpose of this faceted taxonomy is to represent as wide a variety of entities as possible, in order to achieve a truly comprehensive classification of proper names. In addition, and more specific to the present study, this categorization will offer order and terminological value to the list of names extracted from BRIDGET. Table 7.10 below summarises each category in the taxonomy.

175

Hypertypes Types First name Surname Full name Entertainment Figure Cultural Figure Anthroponym Nickname Title Ethonym Zoonym Botanic Celestial Continent Country Region City Village or town Locality Toponym Transportation Hydronym Institution Entertainment premise Retail premise Monument/building Theoretical place Firm Non-Governmental Organisation/charity Artistic ensemble Sport club Chrematonym Public or private institution Political party or organisation Art/literary movement Religion Informal group or association Brand / Product Book /Theatre/Poem Film/TV Song Ergonym Newspaper /Magazine Work of Art Game Individually named object Historical/political event Sporting/cultural event Pragnonym Festival Time period Illness Table 7.10: Ontological facet with its hypertypes and types.

176

Differences with Prolex

One of the main differences between the categorization used by the Prolex group and the adaptation made in this new taxonomy and one of the main reasons why directly applying the Prolex categories to the names from Amanda’s Wedding proved confusing, is represented by the differences between the item levels and the text levels. This refers back to the distinction already pointed out in Chapter 6 of working with types and tokens. As already explained (Chapter 2 & Chapter 6), the distinct proper names occurring in these books are taken from the point of view of the individual referents behind each name. Thus, for instance, a distinction is made between Elvis (the singer) and Elvis (the dog) or between Elizabeth Arden boutique

(toponym/retail premise) and Elizabeth Arden Eight Hour Cream (ergonym/product).

In order to be able to annotate each individual proper name at the item level it is necessary to refer back to the text level and to identify the entity that is being referred to by each name. This distinction at the item and text levels is important as when the results of this classification are combined with the map of translation procedures used to translate each name it will then be possible to identify any pattern in translation procedure usage emerging for each specific type of name, in other words, it could be the case the different procedures are used to translate retail premises to those used to translate product names. This also relates to the polysemy on the text, for instance with the case of the category “institution” which is covered twice in the classification, to refer to two very different uses. Within the hypertype toponym this category refers to education entities where people go in order to acquire knowledge. This is used with verbs of movement, and therefore considered a place and included under toponyms. For instance, the characters in Amanda’s

177

Wedding met at Edinburgh University (toponym/institution). Whereas, within the hypertype chrematonym the category “public or private institution” more widely covers for organisations (including education bodies) founded for a specific purpose or used as an authority in a subject field; the Red Arrows, or the FBI are examples of entities within this type. The group Prolex also has this polysemy distinction to cover for these two types of entities.

Another important difference with the classification by the group Prolex is the distinction made between the different lexical forms that an entity can take. Again, in order to create a comprehensive classification that can be used in other studies and which covers as wide a variety of entities as possible it was necessary to acknowledge this distinction. For instance, it is not rare to find in the names extracted from the corpus that a character is referred to sometimes by their given name and other times by their family name. Although the referent behind each utterance is the same in every case, the lexical difference between them cannot be ignored as it could potentially influence the translation process. This is the reason why the Prolex type “first names” was split into “given name”, “family name” and

“full name”.

Another key difference already pointed out is the introduction of one extra hypertype, not present in the Prolex classification. The hypertype chermatonym was added to categorise the different associations and groups identified in the books.

When attempting to assign a category from the Prolex classification to each one of the names from Amanda’s Wedding, there was continuously a shortcoming with

178 these names referring to groups; in the Prolex classification three types for groups had been placed under the anthroponyms hypertype, and in addition, the type

“firms” was placed under the ergonyms hypertype; however, this felt incorrect as they neither belong to the concrete field alive thing nor are they a product.

Furthermore, there was a number of other groups mentioned that could not fit under the umbrella of any other type or hypertype and ultimately a whole new group took shape. The presence of this hypertype with all its different types gives a new dimension to the faceted classification used with BRIDGET, making it complete and comprehensive.

With the same purpose in mind, other types in the different hypertypes were expanded or refined. For instance, the Prolex categories “celebrities” and “dynasties” were modified. The group celebrities was split into two types “entertainment figure”

(strictly speaking celebrities) and “cultural figure” covering more specifically for writers, scientists, political figures, etc. The entities from “dynasties” in Prolex are covered here. It seems pertinent to mark a distinction between people from the entertainment world and people from the world of culture. In addition, the type

“dynasties” felt too narrow whereas extending the category to cover for all kinds of cultural figures makes it potentially more far reaching and at the same time more accurate to represent a broader subject.

In similar lines, a number of additions were also made into the toponym type. For instance, continent, city and transportation were added. Retail premise and entertainment premise were also added. These locations appear prominently in

179

BRIDGET and represent a common entity that should be covered by a wide ranging taxonomy for proper names. From the point of view of this particular study, again, it opened up the possibility to looking into the relationship of these particular type of toponym and translation procedures. From the point of view of a comprehensive taxonomy, these categories contribute to ensuring the system is a fully comprenshensive and organized schema of subject categories

A number of additions also appear in the ergonyms type, for instance film/tv, song, newspaper, work of art and academic where all added. And the type vessels suggested by Prolex was included into the “individually named object” type.

Although there is no reference to boats in these novels it is important to include it as it could potentially appear in other novels.

The ultimate aim of all these modifications was to produce a classification that would cover for as many entities as possible. This would make the classification more comprehensive, facilitating the replication of the analysis with more Chick Lit books and subject to be used with other kind of studies into the nature of proper names. It is important to point out, however, that as a rule, any type that did not have any figures of use in BRIDGET was removed from the present analysis (in the documents included in the appendixes as well as from Chapter 9) on the basis that if not present in the corpus it could not have any bearing in the translation process of the novels included in BRIDGET.

180

Of course, the subjectivity behind any given categorization cannot be overlooked but the core of these differences between the proposed taxonomy and the Prolex categorization lies in the different aims of the two classifications. The hypertypes and types in this facet attempt to categorise as many entities as possible, offering structure to the data and a terminology to use in the analysis of the procedures. At the same time that attempting to create a taxonomy that could prove usable in other studies. From the point of view of the present study, with this ontological facet, I also sought to organise the data to support the analysis of the translation procedures.

Table 7.4b below repeats the categories included in Prolex (table 7.4a) but with an added column noting the differences with the taxonomy proposed here.

The following sections will now move to describing the other facets in the taxonomy, starting with the semantic facet in section 7.2.2

181

Hypertypes Types Prolex BRIDGET First names Split into first names, given names, full name Celebrities Split between entertainment figure Anthroponyms and cultural figure

Dynasties Included in cultural figure Divinity, mythical or fictive Covered in Sphere facet Firms Covered in new hypertype Chrematonyms

Political Parties Covered in new

hypertype

Chrematonyms

Artistic Ensembles Covered in new Hierarchic Classification -including Sports and music- hypertype Chrematonyms Public or Private Institutions Covered in new -Universities, Hospitals, hypertype Institutions- Chrematonyms NGOs and international Covered in new bodies hypertype Chrematonyms Inhabitants Country Region

Town and villages

Group of countries Toponyms Quarter, Road, Street Locality Buildings -including parks, gardens, theatres and bridges- Hydronyms Geonyms Celestial Objects Fictive or mythical places Private or public institutions Brand name or Trade Mark Firms Moved to Ergonyms chrematonyms Books films, theorems, etc. Mythical object names Vessels Included in new category individually named object Decorations Included in products Meteorological event Historical or political event Pragnonyms Sporting or cultural event Feast name with cyclic character Table 7.4b Prolex classification highlighting main difference with Bridget’s taxonomy

182

7.2.2 Semantic Facet

In the definition adopted for this study (Chapter 2, section 2.3), proper names are seen as functioning both as denoting elements, specifying the one referent, as well as connotative elements, suggesting or implying certain characteristics of this referent. Of course, both are intrinsic to each name, as it is the recognition of the referent which makes the association with its characteristics possible. But sometimes one prevails over the other, and it is this predominance that is being measured in this facet.

This approach takes into account primarily the ‘intended meaning in the mind of the speaker when framing the message’ (Leech, 1981:14) it judges the meaning by the relation of ‘signs to their significates’ (Osgood et al., 1957:3), considering the reason which determines the use of one name or other at one particular moment. The categories in this facet (primarily denotative, intermediate, primarily connotative, described below) have been designed to differentiate between a proper name being used primarily as a referent, e.g. my name is Clara, with the main function of pointing to the name bearer, and a proper name primarily being used as the means to evoke attributes associated with that bearer, e.g. I look like Adam Ant.

This allows for greater systematization in the classification of names, with very clear specific categories following the lines of other facets in the taxonomy. A more quantitative approach to measuring meaning, measuring perhaps the level of connotations of each name or the responses from the readership (as done by

Osgood et al., 1957 or Leppihalme, 1997), would turn the taxonomy into a much

183 more subjective classification, leaving too much to the users’ own perception and imposing his/her ‘own exclusive experience’ which ‘is not necessarily the most reliable source’ (Carter, 1998:209). More in keeping with the aim of this faceted taxonomy for proper names of offering structure to the data at the same time that covering for the aspects of the names more relevant to the translation process, this facet considers exclusively whether an intention of transmitting more ideas than just the name of the object being named can be perceived behind the use of a particular name, and it does not try to measure any reasoning behind this intention or the level of success of each attempt.

Primarily denotative covers those names with the primary function of pointing to a specific individual or thing. For example, the opening statements of Knight’s book

My life on a Plate include the sentence: ‘My name is Clara’ (p.2). This name primarily introduces the character and at this point the reader does not have much information which he/she can associate with this name. The usage is primarily denotative, setting the scene and presenting one of the characters.

Of course, the characteristics which could be associated with a name are always latent in every use, but the main reason for the name in this instance is primarily denotative, and therefore these characteristics become secondary or additional. This can be better visualized in the following example: Fraser Alasdair McConnald is one of the main characters in Colgan’s novel Amanda’s Wedding. This Scottish name has specifically been chosen as the character is Scottish and the novel has an important

Scottish theme running through it. This name, either complete or in the shorter

184 forms of just Fraser or Mr McConnald, appears 170 times within the book, and in all the cases the name is used by other characters when talking to or about Fraser, with a primarily deictic intention. Although the Scottish connotations associated with the name are present in every use, the connection with the character’s name and surname and his nationality is not the primary reason why the name is being used, naming the character is, and therefore this is a primarily denotative name.

Other examples include other characters’ names in the novel, such as Larissa or

Portia, clearly “posh” names as opposed to Melanie or Steve, more “ordinary” names.

On the other hand, primarily connotative covers those names whose main function is to point to the characteristics associated with the referent of the name.

The idea is that recognition of the referent behind the proper name triggers an association which then helps to guide the reaction of the reader. The names are being used primarily for these associations and with the intention to evoke a reaction, and not with a deictic function in mind. Usually these names are embedded in the context in which they are used in the novel and appeal to the reader’s immediate recognition of their referent and the consequent placement of this referent in the source culture. In most cases these names would not make sense without these additional factors. For example, at one point the main female character in Lloyd & Rees’ book Come Together looks at her reflection in the mirror and thinks ‘(…) that I look like Adam Ant’ (p. 274). She has just applied a lot of make-up to try and cover her red and puffy eyes (within the wider context of the novel, the reader knows that she has spent all night crying over her failed

185 relationship). The immediate recognition of Adam Ant triggers the association of this character with make-up (in this case very clearly referring to how he used to paint his face) and the name then acquires meaning. Adam Ant is here being used not as a referential element, to name a person, but for the value of certain connotations associated with the referent. Of course, this association is a stereotypical one which will differ according to class of reader, location, age, culture, and similar values.

Generally, the secret of the success of these novels lies in the ability of the author to choose referents universal enough to be easily recognised by their target readers, and it is in the translation to other languages with different cultures and values where these names acquire a new level of difficulty.

This pattern of behaviour can be best represented in a continuum, where the more extreme usages take the edges, but with the majority of the names resting somewhere in the middle. For ease of representation, this facet contains three categories: the two extremes, primarily denotative and primarily connotative, which have already been described, as well as intermediate, where neither dominates:

P. Denotative Intermediate P. Connotative

Intermediate covers those names where there is not a clear divide, and both the referent and the connotation play a similar role. For example, in the novel Amanda’s

Wedding, Fraser and Amanda met at the Caledonian Ball (p.11), which, again, has very specific Scottish connotations. But when the reader is being told where they met the name is being used as a referential element. However, the characteristics

186 associated with this particular ball, which highlights the upper class, Scottish- aristocracy background of both characters, are equally as important.

For primarily denotative names the connotations are secondary (albeit rarely absent), and can generally be gathered through alternative information present elsewhere in the text; for primarily connotative elements, the connotations are the main reason why the name has been chosen. Failure to understand at least some of these connotations would equal failure to understand what is going on in the novel.

For the intermediate names, connotations play an important role, but they do not affect the main message of the sentence, as the referential function of the name is always clearly present.

Other examples include, for instance, when the specification that the characters drink Absolut vodka, which implies a rather young and trendy clientele, or when they introduce a difference between those characters who go shopping in trendy and expensive Kings Road or Harrods as opposed to those who choose the Elephant and

Castle shopping centre; both specifying where these characters choose to spend their time or money but with clearly very different connotations behind each choice, if the reader sees them.

7.2.3 Scope Facet

A classification based on the scope assesses whether the influence of the name falls only within its context in the novel or, on the contrary, whether it goes beyond the borders of the novel to have a wider meaning in the source culture or further repercussions outside the book. The categories in this facet are:

187

- Internal: referent within the novel

- External referent in the real world

For example, in the novel The Secret Dreamworld of a Shopaholic, the main character is called Rebecca Bloomwood. In this novel, this name is an anthroponym, full name, primarily denotative, with the scope inside the novel. But it may be the case that in another book, Rebecca Bloomwood is referred to through the utterance:

‘you buy more than Rebecca’. In that case, Rebecca would be anthoponym, full name, but primarily connotative with the scope outside the novel. The following case exemplifies the concept more clearly. In the novel Amanda’s Wedding, the two main characters are determined to ruin the wedding and as a first step they plan to ruin the hen party. As they get ready to leave on the night of the celebration they name themselves the ‘Hensterminators’. This name is an anthroponym as it is referring to a person or, in this case, persons. It is a nickname as it is a way of calling those persons other than their birth name. It is primarily connotative, because it has a wider meaning that can only be understood if the reader is familiar with the entities involved in forming the nickname –e.g.: Hen Party + Terminator films-. The scope is however, inside the novel, because the name does not make sense outside its context in this book.

This classification is interesting from the translator’s point of view. As Franco Aixelá

(2000:73) points out, the translator should always consider whether the name has an established translation in the target language. By considering whether the scope

188 of the name lies inside or outside of the novel this facets is taking a similar approach to that which translators have to take.

7.2.4 Sphere Facet

This facet separates fictional names from those that are real in the world and culture of author and reader. It reflects whether or not the referents of the names exist as entities outside the novel. It complements the scope facet and together they map the environment of the name and their relationship with the real world.

There are three categories within this facet:

- Realis – referent real

- Irrealis – referent not real

- Mythical – referring to mythological or spiritual entities (the antichrist,

Heaven, etc.)

The names of characters, both those internal to the novel and those external that happen to be mentioned in the stories, (e.g. Homer Simpson, Mr Darcy, Ally McBeal, etc.) are assigned the category ‘irrealis’. These names are considered irrealis on the basis that, although present in the world outside the novel (and therefore external in the scope facet), they are not real as tangible objects. The combination irrealis/external only applies to names of invented characters from other works of art

189 outside the novel such as books, films, TV programmes, cartoons, even pictures.

This may be more clearly visualized with an example: for instance, the name Doctor

Who, when it refers to the TV series, would be realis and external – as a real entity in the world outside the novel; whereas, on the other hand, the name Doctor Who, referring to the character, would also be external but irrealis – as it does not exist in the real world.

The need to mark this distinction became apparent during the assignation of a category from each facet to each name (see 7.3 below). The benefits from marking such differences within a classification of the names quickly transpire as the most valuable thing to do, as it stresses potential fields which could affect the translation process. With a simple look at the complete classification of the name across all facets it will be possible to view at a glance the various aspects involved in the name, and this is the beauty of a faceted taxonomy; maybe each facet individually comes across as incomplete, but it is in the combination of all the facets where the true nature of each name is depicted. Much richer than uni-dimensional or hierarchic classifications.

7.3 Methodology

A total of 1,610 distinct proper names were identified in the English STs in BRIDGET.

The first approach to the analysis of the data consisted in classifying all the names, with a view to offering a structure to the data and a clear terminology to refer back to during the other stages of the analysis process. The opening sections of this

190 chapter described in detail the taxonomy created to undertake this classification of the data. In the following section I will move to describing the methodology used for applying this taxonomy and present some conclusions from the outcome of the process.

As indicated in the opening statements of this chapter, three Excel documents, one per book in the corpus, collected all the names extracted from each novel. These were arranged in alphabetical order, according to the SL texts. Each of these documents included a sheet (called “class”) with three individual columns, the first one to the left with the English names, the middle one with the corresponding

Spanish translations and the third one to the right with the corresponding Italian translations. Five more columns were added for each language, one per facet in the taxonomy, with the exception of the ontological facet which had two columns in order to cover both for hypertypes and types. The final document has one column with the SL names, followed by five columns for hypertypes, types, semantic facet, sphere facet and scope facet (figure 7.9). A column with the Spanish translations then follows with five more columns for the taxonomy (figure 7.10) and then the same for Italian (figure 7.11), first the Italian names then space for the taxonomy.

191

Figure 7.9: Excerpt from Amanda’s Wedding spreadsheet for classifying SL names.

Figure 7.10: Excerpt from Amanda’s Wedding spreadsheet for classifying TL names – SP.

Figure 7.11: Excerpt from Amanda’s Wedding spreadsheet for classifying TL names – IT.

192

The entire document can be viewed in the appendixes included in the enclosed CD.

The amount of data included in each sheet of these Excel documents make it impossible to fit in a single screenshot or impractical to print legibly.

Pick lists were created for each column with the corresponding categories for each facet, for instance, each cell in the semantic facet column has a drop down menu with three possibilities: primarily denotative, primarily connotative and intermediate.

The same was created under each facet, with their corresponding categories. This facilitates the selection of one category per name under each section. This was replicated for all three languages. Simultaneously, four additional sheets (one per language – named count en, count sp, count it-, plus one more for the totals – named overall count-) were added to the spreadsheet, with an aim to collecting all the quantitative information from the classification performed on the original list.

Conditional and mathematical formulas (COUNT IF and SUM) link each cell and each pick list in the first sheet with the corresponding sheet for each language making it possible to automatically account for every instance of each category. In other words, every time one category from the taxonomy is selected in the pick cell, it is counted for and automatically added to the corresponding cell for the total for that category. In this way it is possible to count the use of each individual type of name, within each language as well as in total. The idea behind this was to observe whether any changes in the category of the names were visible and due to the translation process. Systematically automating the counting makes it less prone to human errors in the count down and allows an easier manipulation of the data.

193

Using the multilingual concordancer, a search for each name in the list was performed. Looking at the individual contexts, a category from each pick list was assigned to each name, both in the originals and the translations.

In the first instance, the ontological facet included the categories extracted from the

Prolex classification. However, as seen in previous sections of this chapter, finding categories to match each type of name in BRIDGET proved difficult and the Prolex classification was consequentially adapted to the specific names in the lists

(explained in section 7.2.2.1). The differences in both classifications have been explored above and illustrated in table 7.4b. Fine tuning of the other facets also took place following a first run through the data and taxonomy, for example, the decision to class all the names of invented characters as irrealis (7.2.4) took place at this stage.

The results from this process remarkably show that no changes to the category of name are carried out during the translation process. Basically, all the same categories are maintained throughout both in SL names and TLs names. Therefore, one conclusion already emerging is that regardless of any translation procedure employed to transfer proper names from SL texts to TL texts, these procedures do not alter the type of name. To a certain extent, this was to be expected within the ontological facet, that is, for instance, the names of the characters would of course, remain anthroponyms in the translations. And the same could be expected to apply to the toponyms, ergonyms, and so on. But it is somewhat more unexpected with regards to the other facets in the taxonomy, primarily the semantic facet. Still, no

194 visible difference was appreciated. It remains to be seen, if the different categories bare any weight with regards to the translation procedures used (Chapter 9).

One last step was taken in order to ensure that this system of categories could be used with a certain degree of reliability (reliability here understood as “the attribute of data to stand in place of phenomena that are distinct, unambiguous, and real”

(Krippendorff, 2008:356)). Content analysis involves testing a given set of categories in order to ensure that these categories are reliable and thus that the results generated by them can be confidently used in other analysis.

An experiment of the inter-annotator agreement was conducted with these data as the means to test its reliability. A sample of 100 names from Amanda’s Wedding were randomly chosen and classified again. The time span between the first classification and the second was considerable, and the original classification was removed while the work was being undertaken, thus ensuring that the new annotation was pure enough to be considered valid for the purpose of this experiment. Scott’s π (Scott, 1955) was used to assess the level of reliability of the data.

In order to calculate this I first established the proportion for each of the facets where the second classification was the same as the first, this is referred to as the observed proportion of matches. In order to calculate Scott’s π from this you also need to assess the proportion that would be expected to be the same by chance.

Scott (Scott, 1955) describes a method of doing this that takes into account the fact

195 that all categories in the facet are not equally likely to be chosen. The formula he proposes is the sum of the squares of the individual proportions of each category. If

Po is the observed proportion that match and Pe is the expected proportion that match then π can be calculated as follows:

π =

The tab ‘Inter-annotator agreement’ in the spreadsheet in Appendix 2 – Amanda’s

Wedding contains the new classification and the calculations required to generate π.

The values obtained in this experiment for π for each facet are as follows:

Facet Π

Hypertype 0.89

Type 0.83

Semantic 0.82

Sphere 0.83

Scope 0.85

Krippendorf (ibid., 2004:241) suggests a value of π greater than 0.8 is required to

‘ensure the data under consideration are at least similarly interpretable by researchers’ (ibid.). All the values obtained are above that threshold therefore the experiment has shown that this system of categories is repeatable, at least by the same researcher over an extended gap period. Thus, the taxonomy has proven

196 trustworthy and the classification can be used with confidence in the rest of the analysis of this thesis.

7.4 Conclusions

The opening sections of this chapter introduce the taxonomy classification devised to structure the data extracted from BRIDGET. This taxonomy takes a faceted form, following recent developments in information science classifications. The categories within each facet were conceived from a combination of both existing classifications

(for instance, that used by the group Prolex in their attempt to compile a multilingual dictionary of proper names), as well as the theories of meaning explored in Chapter

2. Section 7.3 described the methodology employed to apply the taxonomy to the names extracted from BRIDGET and explores its potential and the information that can be inferred from its application to BRIDGET.

Overall, the model and process devised for classifying the data proved both usable and useful. The multilingual concordancer (described in section 6.4.2.1) made it feasible to locate the names in all texts (STs and TTs) at great speed, and the pick lists and formulas automated the quantification of the figures, ensuring an error-free account of the use of each category. The final spread sheet (Appendixes 2, 3, 4), with all the names classified can be manipulated as desired in order to filter the names as needed, making it possible, for instance, to locate and group all the anthroponyms together, or all the connotative, irrealis toponyms, for instance. This offers great flexibility, and ensures clarity and efficiency in the results. All these

197 features will prove useful in the incoming stages of the data analysis, as described in

Chapters 8 and 9 to follow.

Furthermore, the inter-annotator agreement experiment conducted showed that the categories can reliably be used and trusted, thus strengthening the value of this taxonomy.

198

Chapter 8. Data Analysis 2: Classifying the Translation Procedures

“Richard edged over. ‘Um, Door, could you tell the rat something for me?’ The rat turned its head towards him. ‘Miss Whiskers says that if there’s anything you’ve got to say to her you can tell it to her yourself’, said Door. ‘Miss Whiskers?’ Door shrugged. ‘It’s a literal translation’, she said. ‘It sounds better in rat’. Richard did not doubt it”. Gaiman, N.: Neverwhere (p.170)

8.1 Introduction

Chapter 4 of this project included a review of different proposals for the study, classification and categorization of the techniques available to the translator when dealing with culture-specific elements. The review started with Newmark’s (1988) categorization of translation procedures to translate culture, then explored Hermans’

(1988) work with proper names to uncover translational norms and Leppihalme’s

(1997) approach to the study of proper names and other specific elements from a cultural point of view. The chapter also covered Franco Aixelá’s (2000) and Marco’s

(2004) detailed maps of procedures for translating cultural elements. These initiatives all have individual strong points but also some shortcomings. It is the intention of this study to propose a map that combines these strong points at the same time as attempting to mitigate the shortcomings. With the amalgamation of all these projects into one map the aim is to contribute to the study of procedures from the standpoint of previous studies; building on their successes and learning from their faults. It is important, however, to recognise the issues surrounding any given categorization, subjectivity amongst many others. The following sections of this

199 chapter will review the strong points as well as the shortcomings of these models as well as introduce the model adopted here and how it attempts to mitigate these shortcomings and other issues such as subjectivity or arbitrarily. This chapter, thus, illustrates the analytical method used with the names extracted from BRIDGET.

Hermans’ (1988) initial initiative is paramount as it highlighted proper names as key elements for studying translation procedures and norms and succeeded in bringing proper names to the centre of the study of this subject. However, his prime interest was the study of translational norms, and as Hermans points out himself (ibid:24), he left the door open to more work in the area. This challenge was taken up by

Franco Aixelá (2000) who building up on Hermans’ work, presented one of the most exhaustive categorizations of translation procedures for proper names available. His model is ‘based on the degree of intercultural manipulation’ (Franco Aixelá, 1996:60) that the names are subjected to, and the twelve procedures that he identifies are thus divided into two blocks, conservation and substitution, according to more or less adaptation to the target culture. This comprehensive classification has been criticized, however, for lacking a clear delineation between some of the procedures:

‘the contrast between the examples here do not seem very sharp’ (Davies, 2003:71) as well as for its use of a ‘dramatic’ (ibid) terminology somewhat too extreme for the phenomena observed. Davies uses Franco Aixelá’s model as the means to illustrate her ideas around the selection of some procedures over others and how this is not linked to the level of foreignizing or domesticating present in a text. With regards to

Franco Aixelá’s model Davies (ibid.), moreover, also points out an issue with the ordering of Franco Aixelá’s procedures according to the levels of manipulation: ‘some

200 might feel that an extratextual gloss constitutes a further move away from the ST than an unobtrusive intratextual one’ (ibid:71). A similar point has already been illustrated in Chapter 4 of this thesis when discussing Franco Aixelá’s model. It is certain that Franco Aixelá groups the procedures according to their level of adaptation to the target culture into either conservation or substitution, but to my knowledge there is no attempt on his part to then order the procedures within each individual block, so the extent to which this criticism can be strictly applied to Franco

Aixelá’s model is not at all clear.

Such criticism could perhaps be more suited to the continuum proposed by Marco, which is ordered around the level of translator intervention, as it seems arguable that the use of one procedure over another implies more or less intervention on the part of the translator. Although the novelty of placing the procedures on a continuum is welcome here, we agree with Davies that ordering the procedures according to translator intervention or manipulation levels is an approach that can potentially be subjected to critical interpretation. The primary criticism is the main one which can be applied to all taxonomies, that of subjectivity and arbitrariness. In the specific case of the continuum, ordering of the procedures by translator intervention or manipulation is perhaps indeed questionable, however the attempts are commendable and, as Marco points out (2004:133) a continuum seems the most adequate way to illustrate the progression between the procedures. To my mind, what this highlights above all is the need to identify a set of strong criteria upon which to place the procedures on a continuum. And I shall return to this point in

201 section 8.3 below, but before placing the procedures in a continuum, I will start by describing the procedures included in the map proposed in this study.

8.2 Proposal of Map of Procedures

The taxonomy presented in this study consists of a total of 12 procedures, each one illustrating one possible technique to transfer a proper name from SL text to TL text.

The procedures, described in detail below under individual headings, are: borrowing, fixed translation, transcription, literal translation, gloss, explicitation, adaptation, generalization, cultural substitution, cultural replacement, omission and insertion.

Table 8.11 briefly summarises each procedure with one illustrative example (either in Spanish or Italian) from the corpus. The relationships and diversity between the different approaches and the thought processes that lead to selecting one term or procedure over the other possibilities explored in the projects revised in Chapter 4 and summarised in the introduction to this chapter, are also addressed, individually within each heading.

202

Procedures Definition Example

BORROWING Maintain SL name Amanda – Amanda

FIXED Use of established translation in London – Londres TRANSLATION TL for SL name TRANSCRIPTION Adapt SL name to TL Veronica - Verónica morphology, pronunciation, etc. LITERAL Translate SL name word by word Caledonian Ball – el baile de TRANSLATION Caledonia GLOSS (Intra/Extra) Added info (in body of the FHM – la revista FHM text/externally) EXPLICITATION Make explicit in TT what was BP – British Petroleum implicit in ST ADAPTATION SL name is adapted to TL culture Stars in their Eyes – Lluvia de estrellas GENERALIZATION SL name is replaced by CN The Met – la policía CULTURAL SL name is substituted by TL Maris Pipers – coles de Bruselas SUBSTITUTION name CULTURAL SL name is replaced by SL name Anthea Turner – Kate Moss REPLACEMENT OMISSION No name in TT Man U – Ø INSERTION Name in TT but not in ST Ø – Orson Welles

Table 8.11: map of procedures with examples from BRIDGET

8.2.1 Relation to previous maps

In an attempt to disambiguate the terminology, for the classification proposed here we engage with terms from previous procedure taxonomies. For example, like Vinay and Darbelnet (1958/1995) we use borrowing to categorize the use of the name in its original form, or literal translation when the name is translated word-for-word; likewise, we make use of well-known concepts as transcription, explicitation or gloss.

Table 8.12 below shows, in chronological order from left to right, the correlation

203 between the different terms used in previous maps and the terminology proposed here, which is included in the final column to the right.

Chapter 4 explored the relationship of culture-specific elements and translation through a detail description of a range of maps of procedures and approaches to the study of the translation of culture. Taking Newmark’s (1998:94) definition of culture to illustrate the role that proper names play as cultural elements (section 4.3.1.), the proposal presented in this study takes as its main bases the detailed categorization devised by Newmark (ibid.:103). There are, however, certain modifications to this, as some of the procedures envisaged in Newmark’s model are left out (see 8.2.14) and some of the terms are replaced by alternative forms adopted from the other classifications considered. The aim in this new proposal is to harmonize previous maps in order to present a refined but effective taxonomy, with clear terms, combining previous efforts into one approach which could then be applied and replicated with similar studies, be it of proper names in particular or less specifically of any other culture-specific element.

204

FRANCO NEWMARK HERMANS LEPPIHALME MARCO CANTORA AIXELÁ

Transference Copy Keep Repetition Borrowing Borrowing Accepted Terminological Fixed standard Equivalent Adaptation Translation translation Orthographic Neutralisation Transcription Transcription Adaptation Literal Linguistic Literal Literal Translate Translation Translation Translation Translation Keep + guidance Gloss (extra or Gloss Keep + detail Ampliation Gloss intratextual) explanation Omit but transfer Naturalization Neutralization Explicitation sense Cultural Equivalent Ideological Intercultural Adaptation Componential adaptation adaptation Analysis Absolute Classifier Replacement Generalization neutralization Replace by TL Cultural Substitute Naturalization name Substitution Replace by SL Limited Cultural Cultural

name neutralization Adaptation Replacement Non Deletion Omit Omission Omission Omission translation Autonomous Creation Insertion Insertion creation

Label

Compression

Couplet

Table 8.12: Relationship between the procedures in previous models and map used with BRIDGET

The following sections will describe each procedure included in the map proposed in this

study. Each heading below defines one procedure and includes two examples from

BRIDGET one from Spanish TTs (examples (a)), and one from Italian TTs (examples

(b)). For ease of representation the examples are numbered.

205

8.2.2 Borrowing

The name is maintained in the TT; likewise, the referent remains the same.

(a.1) SL Spanish

Amanda Amanda

(b.1) SL Italian

Alex Alex

This procedure is included in all the classifications visited. The term is adopted from

Vinay and Darbelnet’s (1958/1995) original taxonomy for translation procedures and

is also respected by Marco (2007). It corresponds to what Hermans (1998) calls

‘copy’ and Leppihalme (1997) ‘keep’. Franco Aixelá (2000) refers to this technique as

‘repetition’ and Newmark (1988) as ‘transference’.

Borrowing seems a straightforward technique which consists of reproducing in the

TT the name in its original form; however, there are several aspects of this

procedure which deserve further consideration. For example, this procedure has

generally been described as maintaining the SL name. However, it seems pertinent

to consider that it is also used with names which originate in another language, not

the SL, but which are transferred without modifying the spelling in any way. The

examples below, for instance, are names of brands that the characters use: example

(a.2) is the name of a brand of cosmetics and example (b.2) the name of a fashion

label. Both names originate from French, but their form is maintained in all STs and

TTs; similarly the referent is also unmodified:

206

(a.2) SL Spanish

Clinique Clinique

(b.2) SL Italian

Le Coq Fantastique Le Coq Fantastique

This is an interesting point which makes the terms ‘repetition’ used by Franco Aixelá or ‘copy’ used by Hermans seem perhaps less appropriate, as it is not simply a case of repeating or copying a SL name, but rather one of borrowing a foreign term. Let us take into account, for example, the cases where the name is maintained but the form has become fixed, pre-established in the target language:

(a.3) SL Spanish

Margaret Thatcher Margaret Thatcher

(b.3) SL Italian

Nelson Mandela Nelson Mandela

The term borrowing describes the process through which a name becomes acquired into another language.

Similarly, this procedure also covers the cases where the pronunciation of the name varies from language to language, but the graphic form prevails:

(a.4) SL Spanish

China China

(b.4) SL Italian

Sarajevo Sarajevo

207

The key aspect surrounding borrowing is that both graphic form and referent of the

SL name are maintained unchanged in the TTs.

8.2.3 Fixed Translation

The name has a fixed, pre-established form in the TL which is different from that of the SL and which the translator respects. The referent behind the name remains unchanged.

(a.5) SL Spanish

Prince William Príncipe Guillermo

(b.5) SL Italian

China Cina

Again, almost all the taxonomies include this procedure, as it is a basic rule of translation practice:

Al traductor individual la casuística de resolución traductora de los nombres propios le viene ya dada por un hábito linguístico previo de la comunidad en que se haya inmerso, y nada, o muy poco, puede hacer él por modificar ese hábito. Si lo hace, puede incurrir en el más sonoro de los ridículos. (Santoyo, 1988:45)

Newmark (1988) calls this procedure ‘Accepted standard translation’, Franco Aixelá

(2000) ‘terminological adaptation’ and Marco (2007) ‘equivalent’. The main aspect with this procedure is that the graphic form of the name differs from ST and TT.

208

This is an important point, as it marks the difference with the cases of fixed borrowings, where a name enters the TL in the same form it has in the original,

(section 8.2.2).

8.2.4 Transcription

Transcription, another well-known linguistic concept, is used when the graphic form of the name is modified in order to adapt the term to the morphologic or phonologic systems of the TL, for example adapting it to its alphabet or spelling norms. The name is modified but the referent still remains intact. This term is adopted from

Hermans (1988). Newmark (1988) calls this procedure ‘neutralization’ and Franco

Aixelá (2000) ‘orthographic adaptation’.

This procedure is more prominently used with more distant languages, for example

Arabic or Russian where they make use of a different alphabet. With more closely related languages, as is the case of the TLs in BRIDGET, it is more rarely used; it covers cases such as, for instance, added accents in the Spanish TTs:

(a.6) SL Spanish

Veronica Verónica

8.2.5 Literal Translation

Some or all the components of the SL name are translated word for word. The referent does not change.

209

(a.7) SL Spanish

Caledonian Ball El baile de Caledonia

(b.6) SL Italian

Mr Perfect Mister Perfezione

This procedure is included in Vinay and Darbelnet (1958/1995) original taxonomy as well as by Newmark (1988) (‘literal translation’), Hermans (1988) (‘translation’),

Franco Aixelá (2000) (‘linguistic translation’), and Marco (2007) (‘literal translation’).

8.2.6 Gloss

With this procedure the original name is preserved, and therefore the referent is consequently also transferred across to the TT, but with the addition of extra information which serves as a guide to explain, for example, the type of name behind the referent (a.8 and b.7), or describing a reference in more detail (a.9 and b.8).

(a.8) SL Spanish

FHM La revista FHM

(b.7) SL Italian

Waterloo Stazione di Waterloo

210

(a.9) SL Spanish

Abigail’s Party Abigails Party*

* Filme televisivo basado

en una obra de teatro de

los años setenta. (N. de

los T)

(b.8) SL Italian

Agent Provocateur Agent Provocateur*

*catena di negozi di

lingerie femminile

This term is adopted from Newmark’s (1988) taxonomy and is also included by

Leppihalme (1997) (‘keep + guidance’ or ‘keep + detail explanation’), Franco Aixelá

(2000) (‘gloss’) and Marco (2007) (‘ampliation’). An important distinction pointed out by all the scholars refers to the locations of the extra information. When the information is added in the body of the text the gloss is intratextual, whereas if the translator uses a glossary or a footnote, the gloss is then extratextual. The importance of this difference generally lies in the interruption of the reading pace which is generally accompanied by an extratextual gloss. Perhaps not so expected to happen when reading a Chick Lit novel. On the other hand, the information included in an extratextual gloss tends to be more concise and dense than in an unobtrusive intratextual gloss.

211

8.2.7 Explicitation

The translator makes ‘explicit in the target text information that is implicit in the source text’ (Klaudy, 1997:104).

This procedure covers when the translator decides to supplement the TT with added information disambiguating some information that was only implied for the SL readers or information that the SL readership would be expected to know or be familiar with but where the translator judges necessary to explicitly guide the TL readers. Gloss could be considered a mild form of explicitation, as extra information is added to the text in order to make it clearer, more explicit, for the target reader.

However, ‘addition is not the only device of explicitation’ (Séguinot, 1988:108). In some cases the proper name might be omitted from the translation and a whole explanation of its connotations would be offered instead, in other occasions, both name and explanation would cohabit in the TT.

(a.10) SL Spanish

BP British Petroleum

(b.9) SL Italian

Shoestring (-style un giubbotto di pelle e

moustache) due baffetti ridicoli)

Explicitation is a well-studied concept (Olohan and Baker, 2000) (Klaudy, 1997)

(Séguinot, 1988) considered one of the universals of translation (Baker, 1996, see

212 section 1.4). When applied to culture-specific elements, Newmark describes this technique with two procedures ‘naturalization’ or the explanation of the function or external characteristics of the referent and ‘componential analysis’, which sees an added explanation of all the original characteristics of the SL term. Leppihalme

(1997) considers the possibility of transferring the sense but omitting the original name, and Marco (2000) talks of ‘neutralization’ when the function or characteristics of the SL element are overtly explained in the TT.

Before moving on to the next procedure it is important to disambiguate the distinction between gloss and explicitation. Gloss involves the introduction of visible added information, normally only one or two words in the case of intratextual gloss, a brief and simple way to making specific the referent of the name. More examples from the corpus include Tesco/supermercado Tesco or Kit Kat/chocolatina Kit Kat. In the case of extratextual gloss, the footnotes generally include more detailed information regarding maybe the etymology of the term or for instance the author of a book mentioned in the main body of the text.

On the other hand, explicitation is a more subtle way to providing extra information.

It involves more than addition, the explanation offered makes explicit some connotations deemed obvious for the SL readers, or paraphrases the entire sentence in order to convey the meaning of the message rather than the form, and in these cases the proper name is generally omitted. The procedure is different from omission, however, in the sense that the message is still being transferred across, whereas with omission the message, and therefore the referent, are cancelled

213 completely. With explicitation, a reference is clearly visible and the information is still contained in the text but what was implied for the SL readership is clearly exposed for the target language audience.

The most illustrative example of explicitation in the corpus is represented by the

Spanish translation of Knickerless! in Amanda’s Wedding. The whole context of this particular name is explored in detail in the qualitative analysis of names in context included in section 9.4. I will explore the name briefly here as well. Amanda wants to drag the attention of her lover, Nicholas in a busy pub. As a joke with her friend she calls out to him: “Knickerless!”. This is a pun using the homophone blends of the similar pronunciations between the phrase “I am not wearing knickers!” and the name of the man. The Spanish translator makes explicit that pun in the TT where

Melanie shouts “No llevo bragas”, literally “I am not wearing knickers!”.

The extent of the difference between a simple gloss and the full exposure of the connotations behind a name with an explication is clearly visible with this example.

Another example of explicitation, this time involving the cancellation of the proper name in the TT, is the Italian translation of the ethonym Rubenesque. The Italian translator makes explicit the connotation implied by this ethonym by describing the women as “tondi” (“round”). Familiarity with the artist Rubens and the way he depicted women is needed in order to understand the term in the ST; this information is clearly presented for the Italian readers. This cannot be classed as a gloss, as no specific visible information has been added to the text; nonetheless, the meaning and connotations have been made explicit.

214

8.2.8 Adaptation

Adaptation covers the cases where the names undergo some sort of change in order to adapt them to the TL readership. These changes often attempt to make the name more valid in the target culture, for example by changing the referent for its equivalent in the TL (Stars in their Eyes – Lluvia de estrellas) or by replacing a SL nickname with a term from the TL (Hunky Junky - tamarro tabarro). This differs from explicitation, as the translator here introduces a slight change in perspective, opting instead to move the name towards the TL readers rather than a more specific case of deciphering the hidden information for them. The referents behind the names are often, although not always, changed or modified.

(a.11) SL Spanish

Amanda & Fraser Ltd. Amanda & Fraser sociedad limitada

(b.10) SL Italian

Health Secretary Segretario alla salute

Examples (a.11) and (b.10) above both show cases where the names converted into an equivalent concept in the TL. This is an important factor which separates this procedure from the techniques of cultural substitution and cultural replacement which come a bit further along in the continuum. The key issue here is that this change does not affect the foreignness of the novel. It would be strange to have, for example, an English job title in one of our TTs, but similarly it would also be strange to have a Londoner living in a road with a Spanish name, for instance. The

215 translators have to balance the foreign information that they introduce into the TL and that that is adapted (this idea is further developed in Chapter 10). An illustrative example is represented by the levels of adaptation that nicknames tend to undergo; the SL nicknames with semantic meaning are often replaced by TL nicknames with an equivalent semantic meaning or by some which would be used in a similar situation in the target culture, for example Sweet One – Bomboncito – Coccolona.

This is an example extracted from My Life on a Plate and represents one of the many names used by the husband to refer to his wife. Both TTs use a similar nickname from the target culture often used to refer to your lover.

It is also necessary to consider the difference between this procedure when used to translate, for example, names of TV programmes, and fixed translation. In the example mentioned above of Stars in their eyes – Lluvia de estrellas, the translator chooses to use the Spanish version of the programme where people imitate famous artists; it is fundamentally a different programme, not one that has been dubbed from the source culture. This is different to examples of fixed translation used, for instance, with film titles such as The Godfather – el Padrino, where the film is the same in both cultures; that is, the referent remains unchanged. In the case of Stars in their eyes – Lluvia de estrellas, the referent is modified: although the idea is the same, the new programme has a different presenter, different participants, songs, etc. The referent is adapted to the TL culture, it is not a case of cultural substitution as the change does not involve the introduction of a brand new, different referent, but is more a case of an adaptation to the world of the TL readers (for more information on adaptation, please see (Milton and Bandia, 2009), (Raw, 2012)

216

Both Newmark (1988) and Marco (2007) include a technique for the use of an equivalent term in the TL. Franco Aixelá (2000) calls this procedure ‘ideological adaptation’ and goes as far as to say that it is used when a SL concept would be invalid in the TL culture. Maybe this is another case of a use of ‘dramatic’ (Davies,

2003:71) terminology from Franco Aixelá; though maybe not invalid, it can possibly be said that the concept would produce a strange effect in the TL readership and the translator opts instead to bring the term closer to their culture.

8.2.9 Generalization

The specific SL name is changed for a common noun or a word describing the class or group to which the specific proper name belongs.

Newmark (1988) calls this procedure ‘classifier’ and Hermans (1988) describes it as replacing the SL name by a common noun which denotes the functional attributes of the character. Franco Aixelá (2000) stresses that the new element does not show specific characteristics of the SL culture.

(a.12) SL Spanish

The Met La policía

(b.11) SL Italian

Strepsil Pasticca

217

The proper name here goes from the concrete to the general but, as Franco Aixelá points out the new entity ‘cannot be attributed to any particular culture’ (Franco

Aixelá, 1998:37). The specific referent is lost, and instead the new element offers a more general idea, perhaps of the type of name or of its function.

8.2.10 Cultural Replacement

The SL name is replaced by either another name from the SL, or by a different name from another culture but still foreign in the TL culture, perhaps one deemed as better known by the TL readers. The foreignness of the novel is still maintained, although the referent behind the name has been modified.

(a.13) SL Spanish

Benny (from ABBA) Benny Hill

(b.12) SL Italian

Anthea Turner Kate Moss

8.2.11 Cultural Substitution

The SL name is substituted by a TL name. Again, the referent is modified and this change can sometimes affect the foreign flavour of the novel, although this is not always the case.

218

It is important to stress that this change involves a substitution of the SL name by a

TL name but not necessarily by a name from the TL culture as there are also examples of this procedure being used to substitute SL names by the established form in the TL of another SL name (Miriam Margoyles – Nonna Papera). This use is different to the fixed translation or adaptation cases explored above, as the cultural substitution always implies a change in referent, which the other procedures do not always incur.

(a.14) SL Spanish

Maris Pipers Coles de Bruselas

(b.13) SL Italian

Bagpuss Gatto Silvestro

These two procedures, cultural replacement and cultural substitution have been directly adapted from Leppihalme’s (1997) study. Leppihalme observed these changes as very valid techniques often used by Finnish translators and both Franco

Aixelá (2000) and Marco (2007) include at least one of them in their taxonomies.

The main point stressed by these scholars is that the translator seeks a domesticating effect (Venuti, 1995/2008), bringing the name closer to what they perceive as known or familiar for their readership. A change in referent potentially indicates an aim to make the text more easily understandable for the target readers

– it can be presupposed that the new referent is considered by the translator as

219 more valid, better known in the target language culture, although the connotations or associations of the SL name are also modified with this change.

It is important to emphasise the distinction between these two procedures as their effect in the TT is significantly different. With cultural replacement, the name is replaced by another name still foreign in the TL culture, whereas with a cultural substitution the name in fully acquired into the TL culture. Sometimes it involves a change of a referent for another national referent and other times the form of the name is the national one (for instance, in the example of Nonna Papera above). This brings the name closer to the “universo personal” (Franco Aixelá, 1998:43) of the TL readerships. On the other hand, with a cultural replacement the foreignness of the novel remains intact.

8.2.12 Omission

The SL PN is deleted from the TT.

(a.15) SL Spanish

Canderel Ø

(b.14) SL Italian

Man U Ø

220

The source language name, its connotations and all information around it are omitted from the TT altogether. There is no attempt to compensate for the loss or to explain the name in any way. Sometimes whole sentences or paragraphs would disappear from the translation.

Omission is another well-known translation concept and therefore it is included in all the procedural maps.

8.2.13 Insertion

The translator introduces a proper name where one did not exist in the ST.

(a.16) SL Spanish

Ø Dalí

(b.15) SL Italian

Ø The Sun

All maps include this procedure except Newmark and Hermans.

8.2.14 Compression (Marco), Label and Couplet (Newmark)

221

Three procedures appear on the maps explored that are left out in our proposal:

‘compression’ used by Marco (2004) as well as ‘label’ and ‘couplet’ introduced by

Newmark (1988).

Marco’s ‘compression’ refers to information from the ST which is not included in the

TT. In my view omission or adaptation suffice to cover for this technique. Similarly rarely used is the technique described by Newmark as ‘label’, which sees the temporary introduction of a term in the TL until a fixed translation for the SL term gets established in the TL. A use for this procedure can be justified when working with historic translations, but it does not seem to apply with synchronic research as it does not seem possible to know what terms are temporary or will become fixed at a later stage.

More interesting is Newmark’s procedure ‘couplet’, which considers the cases when the translator uses more than one procedure from the map to transfer one given term. For this study the decision was taken, instead, to aim to establish a map where the categories would be mutually exclusive. This is done primarily in order to ease the data presentation. Like with all taxonomies, there is always a certain degree of subjectivity in all categorizations. Although there are cases where selecting solely one procedure from the map is not an unambiguous task, I considered it important to adhere to one procedure per translation instance. This makes it possible to use a variety of measurements for analysing the results and sets clearer boundaries. Therefore, for the cases where there is overlap, the decision was taken to select the procedure that makes the largest contribution to the overall effect of

222 the novel. This can be best visualized with an example from the corpus, for instance, in the Italian version of My Life on a Plate, a glossary is included at the end of the book with definitions for some of the proper names (9.2.6.2). The translator directs the reader to this glossary at the beginning of the novel for clarification on ‘all British terms’ and then uses borrowing throughout the novel for those particular terms. All these names would have been deemed as ‘couplet’ if applying Newmark’s map (as using both borrowing and gloss), but instead they have been classed solely as ‘gloss’ in this study, marking the procedure which, in my opinion, has a bigger effect and best illustrates the processes that are contributing in that particular instance to rendering the text from SL to TL.

This again highlights the main issues surrounding all taxonomies, where decisions have to be taken and adhered to. It is important, however, to always acknowledge the several possibilities at play. Throughout the description of the procedures adopted in this map, I have tried to address the issues encountered as well as to offer an explanation for the final decisions adopted. This includes for example, potential conflicts around fixed borrowings and fixed translation, gloss and explicitation or adaptation and cultural substitutions or replacements. As mentioned above, the necessity of making the map mutually exclusive was paramount and prevailed all throughout the process of creating and applying this map.

8.3 Continuum

223

In a similar way, strong, non-arbitrary criteria were sought for placing the procedures on a continuum. As already highlighted in the introduction of this chapter, there seems to be a recurrent deficiency in identifying a key measurement with which to order the procedures. As previously pointed out, both Franco Aixelá’s and Marco’s attempts are questionable. With Marco’s measure against translator intervention it can be considered that translator intervention happens at all levels, even when a name is subjected to no change, as the translator is clearly intervening when deciding to respect the SL form for the term. Chapter 3 (3.3) reviewed similar ideas from other scholars, most importantly Pym (1992), who very strongly argues how ‘the act of translation’ (Ibid.:75) happens at all levels, even when using borrowing or fixed translations. Likewise, Franco Aixelá’s levels of manipulation do not seem to form a strong boundary for procedures, as Davies (2003:97) points out.

On the other hand, a set of procedures should be amenable to being placed on a continuum, understood as a sequence in which the adjacent elements are not perceptibly different but the extremes are quite distinct (Oxford’s dictionary of

English definition for continuum). From this point of view, a continuum conforms the most illustrative way to represent the progression between the procedures. This brings into perspective once again the necessity of identifying a different set of criteria to collate the procedures on the cline.

Accordingly, a new paradigm is proposed here regarding the ranking of the procedures along the continuum. It draws from the general theories of proper names presented in Chapter 2 and the definition of proper names adopted from them (2.3). This definition sets proper names as elements both with a sense and a

224 referent. The sense refers to the connotations associated with each name and can vary according to different factors, but the referent is always unique to that one particular name being used at one particular time. It is this unambiguous point, the referent, which determines the sequence in our continuum. The key point is whether the procedure employed alters the referent behind the name, that is, whether the new term identifies a new entity; for instance fixed translation adopts the established form for the name in the TL and therefore maintains the same referent, whereas, on the other hand, cultural substitution replaces the SL name for a TL name bringing with it a new entity.

Thus, the quite distinct extremes, which take the names from Franco Aixelá’s two blocks, conservation and substitution, mark the most opposing procedures, the ones which either entirely conserve the original name, e.g. borrowing, which introduces no changes to the referent, or the ones substituting it for a completely different entity, for instance, the cultural substitution mentioned above. The far right end of the substitution extreme is taken by the more radical procedures, omission and insertion. These procedures both affect the referent in the most dramatic way - either by cancelling it completely, or by creating one where there was no space for it on the original. The procedures plotted between these extremes are, as the Oxford dictionary of English definition for continuum suggests, not that radically different, however, they each mark a slight degree of variation to the referent, for instance gloss and explicitation both introduce a description of some kind of the referent, but more extremely generalization sees the referent replaced by the general description of the group to which the specific name belongs. A dividing line is placed across

225

adaptation as the nature of this procedure sees the referent in some cases

maintained but in some others changed. Figure 8.12 below shows a representation

of the continuum as envisaged here:

Cultural Replacement Explicitation Transcription Intratextual Extratextual Cultural Substitution Borrowing Adaptation Insertion Literal transl Fixed Transl Generalitation Gloss Omission

Conservation Substitution

Figure 8.12: Procedures plotted on a continuum

226

Table 8.13 below addresses the specific changes to which the referents of the names are subjected to with each procedure. As the definition for continuum used above states, the two extremes are markedly different but the differences brought about by the procedures in the middle are not so radical. Again acknowledging a certain degree of subjectivity, the place occupied by the procedures in the middle of the continuum is decided on the grounds of the overall perception of their effect on the text; as such a word-for-word translation can be seen as a less dramatic change than a full explicitation of the meaning of a particular term. The differences between both extremes of the continuum are more marked and more easily appreciated, as informed by the variation on the referent.

Procedure Definition Effect Continuum Extreme Borrowing SL term remains no changes in form or referent Fixed trans Adopts TL name changes in form but no changes in for SL term referent

Transcription Adopts TL changes in form but no changes in Conservation graphics for SL referent term Literal Word-for-word changes in form but no changes in Translation translation referent Gloss Added guidance changes in form with extra information added but no changes in referent Explicitation Added changes in form but no changes to information referent Adaptation equivalent effect changes in form and sometimes referent Generalization SL name changes in form and referent replaced by TL common noun Cultural SL changed by changes in form and referent Substitution Substitution SL name Cultural SL changed by changes in form and referent Replacement TL name

Omission No name in TL referent cancelled Insertion Name in TL but new referent introduced not in SL

Table 8.13: Representation of the changes to the referents.

227

Davies (2003) argues against the claim that ‘the procedures identified can be ranked on a scale according to their degree of adaptation’ and she questions ‘the possibility of drawing up such clear-cut scales’ (ibid:70); indeed attempting to place Franco

Aixelá’s procedures in a continuum based on their ‘multicultural adaptation’ certainly proves difficult. With the introduction of the changes to the referent as the main criterion for placing the procedures on a continuum this study attempts to address this issue aiming to offer a more undivided criterion, diluting this problem of unclear scales mentioned by Davies. It has to be acknowledged, of course, that subjectivity is always to a certain degree present in these categorizations. Although whether the referent is the same as it was in the ST or a different one is quite an unambiguous criteria, that is the referent either changes or it does not, the precise order of the procedures along the continuum is open to argument. The extent to which the procedures affect the overall text was used to create the continuum in figure 8.12, but other variations are also possible. Notwithstanding the above comments, acknowledging the role that the referent plays with proper names is of paramount importance, both for the theories of proper names and their translations, as much as for this study. It is indeed this factor, the changes in referent, which will be most important when analysing the results of this investigation.

As set out in the Introduction and in Chapters 6 and 7, this study uses the continuum, as well as numerical measurements to analyse the translations of the proper names in BRIDGET with an aim to extracting tentative generalizations around the tendencies for the translation of proper names from the point of view of Toury’s

228

(1995) ‘adequacy’ or ‘acceptability’ concepts. These concepts link directly to the main two dividing blocks in the continuum presented above, conservation and substitution. This does not imply, however, that the individual procedures have a radical domesticating or a radical foreignizing effect, (as discussed in detail in

Chapter 10) but the hypothesis supported in this study is that considering the overall procedure usage within the text as a whole should point towards the general orientation of the translation, and allow for the articulation of general tendencies and theories.

8.4 Methodology

Chapter 6 introduced the processes used for the identification and extraction of the names in the corpus, and described the Excel document devised for storing and manipulating them. The first stage of the manipulation involved the classification of the names; Chapter 7 described the way in which this Excel document was modified in order to account for each use of the different categories in the taxonomy for proper names. A similar approach was taken with the map of procedures described in this chapter. Using the multilingual concordancer to perform a search for each SL name, a procedure from the map was then assigned to each translation instance

(one to the Spanish translation and one to the Italian translation). A new sheet was added to the original Excel document called Procedures. This sheet (excerpt in figure

8.13) contains five columns: column A has a copy of the SL list of names in alphabetical order, column B includes the corresponding Spanish translations for those names, the procedures used to translate each name are recorded in column C.

229

Columns D and E replicate this structure for the Italian language (column D has the

Italian translations for the original names in A, and column E marks the procedure used in their translation).

Figure 8.13: Excerpt from Amanda’s Wedding spreadsheet for classifying the procedures

In a similar way to the approach taken for classifying the names (Chapter 7), a pick list was created in columns C and E with a drop down menu including each procedure from the map. One procedure was selected for each translation instance from this menu. Another sheet (procedures results) quantifies each time that a procedure is used. Using this data, graphs with the frequency of each procedure were created; in order to clearly illustrate the trends these graphs are placed against the continuum presented above (8.3). The analysis of all this data and the interpretations and generalizations that can be drawn from them are included in the next chapters.

230

8.5 Conclusions

In this chapter the studies on proper names or translation procedures (already introduced in Chapter 4) by Newmark (1988), Hermans (1988) Leppihalme (1997),

Franco Aixelá (2000) and Marco (2007) are incorporated together into one map, created to use with the names from BRIDGET. This chapter includes a detailed exploration of this map and a description of the criteria used to place the procedures against a continuum. The main issues surrounding the classification and the reasons for each decision taken have also been explored in detail. From that standpoint, a classification of all the translation instances identified from the names in the corpus was performed. Section 8.4 describes the methodology and processes used to perform this classification and concludes setting the road for the analysis of all the data collected. This analysis is undertaken in the chapters that follow. Chapter 9 presents the figures obtained and carries out both a quantitative a qualitative analysis. These findings are later discussed in Chapter 10 in light of all the theories explored throughout this study.

231

Chapter 9. Results and Findings

“ ‘The name of Anne Elliot’, said he, ‘has long had an interesting sound to me. Very long has it possessed a charm over my fancy; and, if I dared, I would breathe my wishes that the name might never change’. Austin, J.: Persuasion (p. 147)

9.1 Introduction

This project studies the way in which proper names are translated into different target languages. The approach taken to look at this problem encompasses a number of different phases. Chapter 6 described the methods used to automatically locate all the proper names in SL texts and those employed to identify the translations of those names in the TL texts. Chapter 7 introduced a taxonomy for classifying all these names and the methods used to apply it to the names extracted from BRIDGET. Chapter 8 described the map of translation procedures, set on a continuum, selected to pinpoint the different techniques that translators employ to transfer proper names. The current chapter works with the outputs of all these processes to present an analysis of the translation procedures identified in the

Spanish and Italian versions of the three books in the corpus. Table 9.14 below shows the exact number of distinct proper names extracted from each book in the corpus. The number of translation instances is obtained by adding together the number of distinct proper names in STs, this is multiplied by two, one per TL, plus the number of inserted names in the TTs (this is the names not present in the STs).

In total over 3,000 translation instances have been analysed (as already shown in the Introduction).

232

Book Author Title Names in ST Inserted names in TTs Book 1 Jenny Colgan Amanda’s Wedding 420 distinct PNs 1 SP TT 1 IT TT Book 2 India Knight My Life on a Plate 564 distinct PNs 1 SP TT 0 IT TT Book 3 Helen Fielding Bridget Jones’s Diary 626 distinct PNs 2 SP TT 1 IT TT Total number of translation procedures 3,226 instances analysed Table 9.14 – Books in the corpus with number of translation instances analysed

The approaches taken to investigate all the data involve both qualitative and quantitative methods. Different techniques for corpus processing were combined and integrated together with general statistical measurements for the quantitative analysis of all the data. More manual work was undertaken for the qualitative analysis. In an attempt to present the results in a clear and concise manner this chapter is divided into three main parts: A): Translation procedure usage – Raw and relative frequency; B): Procedures and taxonomy for proper names; C): Examples of names in context. With a summary of the findings and closing remarks located towards the end of the chapter (9.5).

Part A) in section 9.2 explores the frequency of use of each translation procedure in the continuum. The results are presented twofold, firstly working with the total number of occurrences of each procedure across all TTs and secondly, looking into the figures for each individual book.

Starting the analysis by presenting the figures for the entire corpus gives a first broad idea of the general tendencies that will then be dissected in more detail in the other sections in the chapter.

233

In this section I also review (9.2.2.1) the procedures which turn out to be less employed by the translators, with examples from the corpus as a way to understand the reasons why they might be less favoured. The most popular procedures are analysed in close detail throughout the chapter.

The investigation of the figures by book starts with a quantitative analysis of the data extracted from the raw frequency and percentages of usage of each procedure in the different TTs, Spanish and Italian. Overall intra- and interlinguistic tendencies are examined. Where these highlight prominent patterns they are explored in greater depth with a more extensive analysis.

Finally, section 9.2.6 takes a look at the overall tendencies, across all three books, by language.

In Part B) in section 9.3 I try to determine any possible relationship between name type and procedure usage by looking closely at the translation procedures employed within each individual category in the taxonomy for proper names presented in

Chapter 7. Again, raw and relative frequency measurements are used. The figures are compared with those obtained in part A). The results are presented by category within each facet. The data encompasses all three books and is summarized by language. In keeping with the approach taken in part A), a quantitative analysis is undertaken first, with examples and detailed explanations for each facet; special emphasis is given where variation in the data is clearly visible.

234

Part C) in section 9.4 presents a qualitative analysis of a sample of names in their context, considering the different approaches taken by the translators in the different versions of the novels included in the corpus.

9.2 Translation procedure usage - raw and relative frequency

9.2.1 Introduction

An initial approach to the quantitative data explores the frequency of use of the procedures in the continuum. Raw and relative frequencies are techniques used for corpus processing. Raw frequency counts the number of times a specific feature, in this case a procedure, occurs whereas its relative frequency shows the overall rank of that number in relation to the total in the corpus. All the data extracted from

BRIDGET is presented according to its raw and relative frequency.

Using maths functions in Microsoft Excel, the number of times a single procedure from our map was assigned to a translation instance is counted and the procedures are then ranked according to their popularity. Plotting this data into the continuum illustrates the approach taken by the different translators – highlighting a tendency to either conserve, if the majority of the procedures used are found to fall near the left hand extreme of the continuum, or to substitute, if the majority fall near the right hand one.

The section opens with an overview of the data as a whole – with the relative frequency of each procedure in the entire corpus (this is, adding together all the

235 instances of each procedure across all TTs). This brief introduction points towards the general tendencies in the translation of proper names in Chick Lit which can be discerned.

The findings are then presented by book, starting with the first book in the corpus,

Amanda’s Wedding, and following with My Life on a Plate and finishing with Bridget

Jones’s Diary. An overall summary by language is included at the end of the section.

Within each section, firstly a table shows all the data relevant to the section, that is, the figures for the entire corpus in the overview (9.2.2) and the figures for each individual book in the following sections. The first two columns remain the same, and act as a reference, with the procedures –with example- (column 2) ordered by the place they take in the continuum (column 1). Next, the specific data for each book is given, by language, first Spanish and then Italian. Raw frequency is first and relative frequency, as a percentage, second.

The figures are then analysed, working with the numbers in order to unfold the patterns followed by the different translators. A graph is then presented, plotting the data onto the continuum as a means to clearly illustrate the findings. For My Life on a Plate and Bridget Jones’s Diary the information is put in parallel with that codified for the previous books.

Finally, section 9.2.6 summarizes all the findings by language, across all three books, concretely setting out and separating the tendencies identified for the Spanish language to those identified for the Italian language.

236

9.2.2 Overview

In this first pass through the data, I have added together the number of occurrences of each procedure in the continuum in all TTs (both Spanish TTs and Italian TTs).

Table 8 below collects all the data, by book and as a total, showing the times that each procedure can be accounted for in each TT as a percentage of the total of instances (its relative frequency for the corpus). The last column adds together all the figures for all books, presenting a total percentage for the entire corpus.

With this first segmentation of the data I have sought to obtain an initial idea of the patterns for the translation of proper names in Chick Lit from a general point of view, without variables such as language or book. On the other hand, this approach also contributed to bringing into sharper focus the map of procedures as a whole, as the data clearly points towards some procedures (specifically transcription and insertion) which are very rarely used in all the TTs and which therefore, perhaps call for a new consideration around their role in the continuum. This possibility is explored in closer detail in section 9.2.2.1 with a qualitative analysis of these procedures in context.

237

The data: Overview

My Life Bridget Amanda’s Continuum Procedure on a Jones’s Total Wedding Plate Diary

Conservation Borrowing Amanda – Amanda 55.34% 54.93% 52.83% 54.22% Fixed Translation London – Londres 13.06% 14.12% 20.35% 16.28% Transcription Veronica – Verónica 0.12% 0.09% 0.00% 0.06% Literal Translation Caledonian Ball – el baile de Caledonia 9.98% 9.86% 9.10% 9.59% Gloss (intra/Extra) Ryvitas – Galletas Ryvitas 3.21% 8.87% 0.88% 4.25% Explicitation BP-British Petroleum 2.02% 0.45% 1.04% 1.09% Adaptation Stars in their Eyes – Lluvia de estrellas 3.21% 3.71% 2.71% 3.19% Generalization The Met – La policía 6.29% 3.71% 7.34% 5.81% Cultural substitution Bagpuss – Gatto Silvestro 0.71% 0.81% 2.08% 1.28% Cultural replacement Anthea Turner – Kate Moss 1.54% 0.27% 0.88% 0.84% Omission Man U – ø 4.28% 3.08% 2.55% 3.19% Substitution Insertion Ø – Orson Welles 0.24% 0.09% 0.24% 0.19% Table 9.15 – Procedure relative frequency across all TTs

The picture represented is initially surprisingly similar for all three books on all the procedures and indeed it seems to reflect a very similar approach from all translators in both languages across all three books. A closer look at the data, from different perspectives, as undertaken below, shows that there are, however, important

238 differences at all levels, book, translator and language. Sections, 2, 3, and 4 below will deconstruct all the data and unfold all the patterns.

Notwithstanding the differences, certain tendencies already begin to emerge from this data as it very clearly shows a very strong predominance of borrowing over all the other procedures in the continuum. By the same token, there are certain procedures, around both extremes of the continuum, that seem to be more popular than others (namely, fixed and literal translation, gloss and generalization over explicitation, cultural substitution and cultural replacement), as well as some procedures barely used at all (transcription and insertion). Examples and detailed analysis of the most popular procedures follow in the next sections.

As just mentioned, the second most striking feature exposed on this overview of the data is the marked low frequency of transcription (0.06%, two occurrences) and insertion (0.19%, six occurrences). These low numbers call for a more closed analysis of these lesser used procedures from a more qualitative point of view, exploring their uses and the factors at play behind and around them in order to try and understand the role they play in the novels as well as in the continuum and map of procedures as a whole.

239

9.2.2.1 – Lesser used procedures

9.2.2.1.1 Transcription

With this procedure the translator adapts the name to the alphabet or spelling norms of the TL. There are only two examples of this procedure in the entire corpus, both in Spanish TTs – one in Amanda’s Wedding and one in My Life on a Plate, and both involve accents:

(a) SL Spanish

Veronica Verónica

The first example is an anthroponym/first name and it appears in Amanda’s

Wedding. This is the name of a work colleague of the main character, Melanie, it has only one token. The Spanish translator adds a tilde on the second vowel, following the grammatical rules on the Spanish language. This could, of course, also be classed as a fixed translation, as this is the common form for this first name in

Spanish.

(a) SL Spanish

Angélina Angelina

The second example is a toponym/entertainment premise and it is from My Life on a

Plate. This is the name of a little café in Paris where Kate, the main character, goes for lunch during a romantic weekend with her husband in France. This is a very clear

240 example of transcription as the name is changed from the original French spelling to the traditional Spanish spelling. The strength in Angelina, from a Spanish point of view, falls on the second syllable, palabra llana, and ends with a vowel so therefore no accent is required. This name has clearly been adapted to the rules of the

Spanish language.

The Italian TTs record borrowings for both these names.

Although the proximity of the languages of this study make it seem almost redundant to have this procedure - as evidenced by the scarce number of examples and the possibility to even include one of them under a different procedure altogether - there is still a strong case for maintaining it in the map, as it would be essential in works with more and/or diverse TLs, such as Russian or Chinese. This procedure is consistently included in all the studies and descriptions of translation procedures and takes an important role where different alphabets are at play (see for instance (Espinal, 1989-1991, also in 3.2)).

9.2.2.1.2 Insertion

This is the procedure where the translator decides to add a proper name where one did not exist in the ST. This is a rare move from the translators, when they decide that the best way to transfer the original to their target audience is by introducing a new referent to the text. There are only six occurrences of insertion in BRIDGET, but nonetheless this procedure cannot be overlooked as is a very valid option for the

241 translators and one that very clearly cannot fall within any other procedure or be classed as any other technique. There are examples in all three books, 4 in Spanish

TTs and 2 in Italian TTs; interestingly, as described below in examples 5 and 6., on one occasion both translators decide to make an insertion at the same point of the novel, adding exactly the same information. It is difficult to assess whether the translators might have been aware of each other’s work, but in any case, it seems relevant that both deemed it necessary to add the same information to transfer the content of this proper name to their respective target audiences.

In order to understand the effects that the translators might be trying to recreate by the use of this procedure a bit of the context of the moment in which each name is introduced is needed and therefore each example is analysed on its own in detail below:

1. La boda de Amanda – Insertion: Orson Welles

(a) SL Spanish

My angel in the snow. Striding like Mi ángel de la nieve. Alejándose a

Mr Rochester from the church — I grandes zancadas de la iglesia como

hadn't actually seen him do that, but el señor Rochester en la película

I imagined it that way. de Orson Welles; en realidad no lo

había visto irse, pero me lo imaginé

así.

242

In this first occurrence of this procedure, the Spanish translator introduces a new referent as an attempt to disambiguate the referent behind the name just mentioned, Mr. Rochester. The action for this name takes places at the end of the novel, when Melanie has decided to leave the wedding scene and is on a train on her way back to London. Bored, her mind starts to drift off and sets off to thinking of Fraser, as she imagines him leaving the church ‘like Mr. Rochester’.

The Spanish version of the novel adds ‘como el señor Rochester en la película de

Orson Welles’, setting a clear referent behind Mr. Rochester, that of the actor directed by Welles in his 1943 version of Jane Eyre.

Trying to speculate around possible reasons for this insertion, one could consider that either the translator was more familiar with the film than with Charlotte Bronte’s novel, or rather that he deemed it easier for his audience to understand the connotations behind the manner in which Melanie was imagining her hero Fraser leaving the church through setting a clear real referent for them to visualized.

Interestingly, the Italian translator omits all reference to Mr. Rochester, and Melanie just dwells on the marvels of her Fraser:

243

(b)

SL Italian

My angel in the snow. Striding like Il mio angelo nella neve. Il suo

Mr. Rochester from the church — I sorriso svagato, il suo sguardo

hadn't actually seen him do that, but gentile, la sua risata...

I imagined it that way.

2. Non sposate quella donna! – Insertion: El Sun

(b) SL Italian

Dreaming of the day Linda Lusardi "A sognare del giorno in cui la

comes past and accidentally breaks ragazza della terza pagina del Sun

down in front of your little Cockney sarebbe rimasta in panne proprio

house.' He held up his arms and davanti a casa tua ". Lui alzò le

walked off. 'I don't have to listen to braccia e se ne andò. "Non sono

this’. obbligato a starti a sentire ".

This example, extracted from the Italian translation, sees an insertion used as the means to specify a referent from the original that the translator might deem as unknown in the TL culture. In this case Linda Lusardi is replaced by a more general referent ‘la ragazza della terza pagina del Sun’, where the specificity of the original is lost. By this move, the translator is also inserting a new name, with a new referent into the TL, the newspaper where the SL referent often appears.

244

The sexual implications behind the original referent are still intact in the translation, as it can be presupposed that the Italian audience is familiar with the content of page three of British tabloids like The Sun, but perhaps not to the level of knowing the models by their names.

The Spanish translator simply transfers the name across with a borrowing, without modifying the referent in any way.

3. Mi vida sobre un plato – Insertion: Los Baker Boys

(a)

SL Spanish

I scrumple him up, in fact, and shred Lo arrugo, de hecho, y lo hago trizas

him in the Magimix, razor-blade en el Magimix, con sus pómulos

cheekbones and all. Nobody nice is afilados y todo. Nadie que sea

that cartoonishly good-looking and simpático puede tener esas facciones

he'll make lovely fresh bedding for de dibujo animado, y se convertirá en

Binky and George Roid (the Hammy un lecho nuevo para Binky y George

Roids, named, I'm afraid, by mature Roid (los Hammy Roids, como los

me). Baker Boys).

At this point in the novel My Life on a Plate, the main character, Clara is distractedly trying to prepare an interview with a celebrity modern dancer that she has to conduct for the paper where she works. The dancer has agreed to visit her in her

245 home and while she waits she busies herself with different housework chorus. At one point she takes to using his CV and information report as bedding for the hamsters, called by her the Hammy Roids. This is a pun, combining the sounds of

Hammy for hamster and roid for rodent to sound like haemorrhoids, which the reader presumes describes how these animals make her feel.

The Spanish translator maintains the name that Clara has given the hamsters,

Hammy Roids, but adds a new referent like the Baker Boys. The pun of the original, is lost for an audience unfamiliar with English phonetics, although a rhyme is maintained between roids and boys. The Baker Boys can possibly refer to the 1989 film where two artist brothers see their career revamp by hiring a new female singer. This could perhaps refer to the profession of the celebrity whose CV now serves as a bed for the hamsters (a modern dancer). The exact purpose behind this insertion is not known.

The Italian translation takes a very different move, adapting the name to the Italian language, where Cricetini is the word for hamster while maintaining their surname

Roid. The joke and connotations of the original are somehow lost.

(b)

SL Italian

I scrumple him up, in fact, and shred Anzi, in realtà, lo appallottolo senza

him in the Magimix, razor-blade tanti riguardi e lo spezzetto nel

cheekbones and all. Nobody nice is Magimix, zigomi cesellati e tutto.

that cartoonishly good-looking and Nessuna persona come si deve

246

he'll make lovely fresh bedding for possiede questa bellezza da cartone

Binky and George Roid (the Hammy animato, e in ogni caso lui è destinato

Roids, named, I'm afraid, by mature a diventare la deliziosa, nuova lettiera

me). di Binky e George Roid (i Signori

Cricetini Roid, così battezzati, temo,

dalla matura sottoscritta).

4. El diario de Bridget Jones – Insertion: Dalí

(a)

SL Spanish

Sometimes I wonder what I would be A veces me pregunto cómo sería

like if left to revert to nature — with a volver a la naturaleza: con barba y

full beard and handlebar moustache con un bigote estilo Dalí

Bridget is getting ready for a date and as she exhausts herself with all the beauty preparations needed she considers what she would look like if left ‘to revert to nature’. Amongst the many things that would grow she envisaged a ‘handlebar moustache’. The Spanish translator introduces a reference to the Spanish artist Dalí to describe the kind of moustache Bridget imagines. The image of Dalí will immediately bring a twisted moustache to the head of any Spanish reader, so this insertion serves to recreate the connotations of the original perfectly.

The Italian translator does not include any referent, and simply, like the original, uses the conventional term for this kind of moustache.

247

The last example of this procedure at use involves an insertion of the same term in both TTs, Spanish and Italian:

5. El diario de Bridget Jones – Insertion: Orgullo y prejuicio

(el de) Cumbres Borrascosas

(a)

SL Spanish

It struck me as pretty ridiculous to be Me pareció bastante ridículo llamarse

called Mr Darcy and to stand on your míster Darcy como el de Orgullo y

own looking snooty at a Party. It's- prejuicio, y permanecer a solas con

like being called Heathcliff and aires de superioridad en una fiesta.

insisting on spending the entire Como llamarse Heathcliff el de

evening in the garden, shouting Cumbres borrascosas e insistir en

'Cathy' and banging your head pasar toda la noche en el jardín,

against a tree. gritando «Cathy» y golpeándose la

cabeza contra un árbol.

6. Il diario di Bridget Jones – Insertion: Cime Tempestose

(b)

SL Italian

It struck me as pretty Chiamarsi Darcy e starsene tutto solo con

ridiculous to be called Mr Darcy aria sdegnosa a una Testa mi ha subito

and to stand on your own colpita come una cosa abbastanza ridicola,

248

looking snooty at a Party. It's- un po' come se, in Cime tempestose

like being called Heathcliff and Heathcliff passasse tutta la serata in giardino

insisting on spending the entire a gridare "Cathy" e a sbattere la testa contro

evening in the garden, un tronco.

shouting 'Cathy' and banging

your head against a tree.

Here Fielding is using an allegory to compare the character, Mr Darcy, to the well- known protagonist of Austin’s novel, Pride and Prejudice. As indicated in section

5.3.3 Fielding took inspiration from this classic novel for the characters and story-line of the book, this comparison here directly links both stories. She is compering the character of Austin’s Mr Darcy, famous for being distant, closed and prejudiced, to the character of Bridget’s friend. The simile is extended to Bronte’s character

Heathcliff, to complete the joke. The connotations of the implications of these comparisons of the characters in Bridget with the characters on these famous classics imply the familiarity of the readers with their referent and what they have come to represent. The translators from both TLs deem it necessary to specify the title of the novel where Heathcliff comes from, Wuthering Heights. As already pointed out, it seems interesting that both translators decide to add the same information at the same point in the novel. In contrast, example 5(a) from the

Spanish text, contains two insertions, as the translator also specifies that Mr Darcy belongs to Pride and Prejudice. Again, it is impossible to presuppose the exact intention of these insertions, but it can be implied that the Spanish translator

249 considered it appropriate to offer this extra piece of information, in addition to supplementing Heathcliff, to his readers.

This more qualitative inspection of transcription and insertion show the roles that both these procedures play on the novels. Although the quantitative analysis of their occurrences throughout shows very low frequencies for them, the more detailed exploration of their uses demonstrates that they are very valid techniques that translators sometimes employ and their place in the continuum cannot be taken by any other procedure.

9.2.2.2 Conclusions

In summary, this first encounter with the data as a whole, and the analysis from qualitative and quantitative points of view, already indicates that certain procedures

(primarily borrowing) are perhaps more commonly used than others; in the next section I will explore the figures for each book individually, comparing and contrasting the languages. This should point to any possible differences with these initial figures taking the data as a whole and will start to indicate possible variations as well as patterns.

250

9.2.3 Book 1

SL Spanish Italian Amanda’s Wedding La boda de Amanda Non Sposate Quella Donna!

The data

Spanish Italian Raw Relative Raw Relative Continuum Procedure Frequency Frequency Frequency Frequency Borrowing Conservation 252 59.86% 214 50.83% Amanda – Amanda Fixed Translation 54 12.83% 56 13.30% London - Londres Transcription 1 0.24% 0 0.00% Veronica - Verónica Literal Translation Caledonian Ball – el 49 11.64% 35 8.31% baile de Caledonia Gloss (intra/Extra) Ryvitas – Galletas 21 4.99% 6 1.43% Ryvitas Explicitation 9 2.14% 8 1.90% BP-British Petroleum Adaptation Stars in their Eyes – 14 3.33% 13 3.09% Lluvia de estrellas Generalization 14 3.33% 39 9.26% The Met – La policía Cultural substitution 1 0.24% 5 1.19% Bagpuss – Gatto Silvestro Cultural replacement 1 0.24% 12 2.85% Anthea Turner – Kate Moss Omission 4 0.95% 32 7.60% Man U – ø Insertion Substitution Ø – Orson Welles 1 0.24% 1 0.24%

Table 9.16: Breakdown of procedures used in each TT – Amanda’s Wedding

From this data, borrowing immediately stands out as by far the most prominent procedure used in Amanda’s Wedding in both languages, with a relative frequency of

251

51% in Italian and of almost 60% in Spanish. Similarly, fixed translation ranks second in both languages, however, a substantial difference appears with the procedures ranking in third place. The Spanish language remains near the conservation extreme of the continuum, with literal translation showing a relative frequency of 12%, but the Italian language moves nearer to the substitution extreme of the continuum with generalization as the third most popular procedure, used in 9% of all the translation instances. Generalization ranks fifth in the Spanish language, matched with adaptation, both with a relative frequency of 3%.

Over 8% of the translation instances in the Italian text are translated with literal translation, a much lower percentage than that identified in the Spanish language.

As generalization showed the strongest difference between the languages a closer exploration of its uses reveals that out of the 39 generalizations present in Italian, eight correspond to generalizations in Spanish. The remaining six Spanish generalizations are translated with a range of different procedures in Italian where no pattern can be drawn.

The other Italian generalizations are also translated with a range of procedures into

Spanish, but the vast majority, 17, are translated with borrowings in Spanish, almost all of these names are classed as ergonyms in the taxonomy.

Another visible difference between the languages occurs where the Spanish translator makes more use of gloss than the Italian one. Brief explanations on the text or a longer one as a footnote are introduced in 5% of the translation instances in Spanish (21 instances). A gloss is only used with 6 names in Italian, only 1% relative frequency, and therefore perhaps not reliable enough to offer any relevant

252 information. In any case, for every instance in the Italian text, the information is added to the main body of the text, none appear as footnotes or other paratextual information. On the other hand, the Spanish text includes a total of 12 extratextual glosses, with the information added in a footnote and marked as ‘nota del traductor’

Around the middle of the continuum, adaptation and explicitation both see similar percentages of usage in both languages, 3% for adaptation and 2% for explicitation, again indicating a modest use of these procedures consistent in both TTs.

The differences in the numbers are greater for the rest of the procedures with consistently higher frequency for the procedures around the substitution extreme of the continuum in the Italian language than in the Spanish language.

The greatest difference appears with omission, hardly used in Spanish, with a relative frequency of under 1%, but showing 8% in Italian. Entire sentences and paragraphs, up to 28, are omitted from the Italian TT with no attempt to compensate for the loss in any way. The majority of the Italian omissions correspond to borrowings or gloss in Spanish.

Finally, a strong difference between the languages also happens around cultural substitution and cultural replacement. These procedures each occur just once in the

Spanish text but show a relative frequency of 1% (five cases) and 3% (12 cases) in

Italian. The majority of the names substituted or replaced in the Italian are anthroponyms, where some celebrity or external character is changed for another one, perhaps better known in the target culture. A more qualitative consideration of

253

this finding is given in 9.4.1 where I look into the specific procedures employed with

anthroponyms.

All in all, this data points towards a greater degree of variation in the Italian text,

with fewer shifts occurring in the Spanish version. This can be appreciated very

clearly from the graph below (figure 9.14), where the bars are higher for the

Spanish language than for the Italian language around the procedures located near

the left hand side, the conservation extreme, whereas they are all higher for Italian

than Spanish around the right hand side, the substitution extreme:

Conservation Substitution

Figure 9.14: Procedure relative frequency against continuum – Amanda’s Wedding

254

9.2.4 Book 2

SL Spanish Italian My Life on a Plate Mi vida casi perfecta La mia vita su un piatto

The data

Spanish Italian Continuum Procedure Raw Relative Raw Relative

Frequency Frequency Frequency Frequency Conservation Borrowing 323 58.41% 284 51.45% Amanda – Amanda Fixed Translation 78 14.10% 78 14.13% London - Londres Transcription 1 0.18% 0 0.00% Veronica - Verónica Literal Translation Caledonian Ball – el 52 9.40% 57 10.33% baile de Caledonia Gloss (intra/Extra) 17 3.07% 81 14.67% Ryvitas – Galletas Ryvitas Explicitation 2 0.36% 3 0.54% BP-British Petroleum Adaptation Stars in their Eyes – 18 3.25% 23 4.17% Lluvia de estrellas Generalization 27 4.88% 14 2.54% The Met – La policía Cultural substitution 5 0.90% 4 0.72% Bagpuss – Gatto Silvestro Cultural replacement 1 0.18% 2 0.36% Anthea Turner – Kate Moss Omission 28 5.06% 6 1.09% Man U – ø Substitution Insertion 1 0.18% 0 0.00% Ø – Orson Welles Table 9.17: Breakdown of procedures used in each TT – My Life on a Plate

From this data borrowing also stands out as the most prominent procedure in My

Life on a Plate in both languages. More than half of all the cases are translated with

255 this procedure both in Spanish (58%) and Italian (51%). Noticeably, however, the relative frequency is considerably higher for the Spanish language at 58%.

In Spanish, fixed translation again ranks second whereas in Italian this procedure ranks third, very slightly behind gloss. The relative frequency of fixed translation is considerably lower than that of borrowing, amounting to under 15% of all translation instances in both languages.

Gloss ranks fifth in Spanish, showing a marked difference with Italian. This difference can be explained by the fact that the Italian version of this book includes a glossary at the end with an explanation of those terms ‘belonging to the British culture’.

The existence of this glossary has already been mentioned in section 8.2.14. and its affect to the data can now be clearly appreciated. Of the 81 cases of gloss that can be found in the Italian version of this book the vast majority, 74, are extratextual. Of the other seven, four are footnotes and the other three are examples of intralingual gloss. Overall, this procedure has a relative frequency of 15% in the Italian TT, similar percentage to fixed translation.

Gloss is only used to translate 3% of all the instances in Spanish, with 17 tokens; of these, three are intratextual with the other 14 introducing the explanation as a footnote.

256

Following on with the analysis of the figures, literal translation shows a low relative frequency, just over 9% in the Spanish TT and over 10% in the Italian TT.

Adaptation ranks fifth in the Italian language and sixth in the Spanish, with a relative frequency of 4% and 3% respectively. The majority of the occurrences of adaptation in both languages are examples of this procedure used with nicknames, where the nickname is changed by a TL nickname familiar for the target readers. The majority of the examples correspond in both languages.

(a) SL Spanish

Fluffkins Ricura

Bunny Pants Gran Dama

(b) SL Italian

Fluffkins Morbidona

Bunny Pants Patatina

These are examples of the terms of endearment that the husband gives to his wife, and it is expected that these would be adapted to the names more commonly used in these situations in the TLs. A more detailed analysis of these nicknames can be found on section 9.3.2 which combines procedures with ontological facet in the taxonomy.

Overall, the relative frequency of the procedures nearing the substitution extreme of the continuum is considerably lower for both languages than that identified for the

257 procedures around the opposite extreme (approximately, 90% of all the translation instances are covered with procedures around the conservation extreme). It is within the procedures around the substitution extreme, however, were the strongest differences between the languages can be seen. Omission, for example, is used in

5% of the cases in Spanish but only in 1% in Italian. Similarly, generalization is used with over 5% of the names in Spanish but with less than 3% in Italian. This indicates a stronger tendency to manipulate the text in the Spanish version of the novel than in the Italian, although, of course, the percentages of use of these procedures are comparatively small in both languages, with an overall domination of the procedures at the conservation extreme of the continuum.

Cultural substitution is used in just 1% of the cases in both languages, with a low raw frequency of 5 instances in Spanish and 4 instances in Italian. Lastly, with an even lower raw frequency (two or less) come insertion and cultural replacement, hardly used in either translation, neither procedure amounting to more than 1% in either language.

Plotting this data in a graph (figure 9.15) sets out clearly the oscillation in the usage of procedures by language, at the same time that it illustrates the marked use of the procedures nearing the conservation extreme in both languages.

258

Conservatio Substitution n

Figure 9.15: Procedure relative frequency against continuum – My Life on a Plate

In summary, the feature of this book is presented by the surge in the use of gloss in

the Italian language. This generally lowers the numbers for all the other procedures

in this text. Apart from that, the general trend is the domination of borrowing over

the other procedures and the similarities between the languages in the numbers of

fixed translations, literal translations and explicitations used. This goes in line with

findings for Amanda’s Wedding, where borrowing and fixed translation also

dominated over the other procedures. However, Amanda’s Wedding presented a

clear difference in the procedures plotted against the continuum, with the Italian

language showing more use of the procedures around the substitution extreme than

the Spanish. This seems to have reversed in book 2, where the Spanish version

seems to present a wider range of procedures used overall.

259

9.2.5 Book 3

SL Spanish Italian Bridget Jones Diary El diario de Bridget Jones Il diario di Bridget Jones

The data

Spanish Italian Continuum Raw Relative Raw Relative Procedure Frequency Frequency Frequency Frequency Borrowing Conservation 351 55.89% 311 49.68% Amanda – Amanda Fixed Translation 131 20.86% 124 19.81% London – Londres Transcription 0 0.00% 0 0.00% Veronica – Verónica Literal Translation 71 11.32% 43 6.87% Caledonian Ball – el baile de Caledonia Gloss (intra/Extra) 8 1.28% 3 0.48% Ryvitas – Galletas Ryvitas Explicitation BP-British 7 1.12% 6 0.96% Petroleum Adaptation Stars in their Eyes 11 1.75% 23 3.67% – Lluvia de estrellas Generalization The Met – La 28 4.47% 64 10.22% policía Cultural substitution 10 1.59% 16 2.56% Bagpuss – Gatto Silvestro Cultural replacement 2 0.32% 9 1.44% Anthea Turner – Kate Moss Omission 6 0.96% 26 4.15% Man U – ø Insertion 2 0.32% 1 0.16% Substitution Ø – Orson Welles

Table 9.18: Breakdown of procedures used in each TT – Bridget Jones’s Diary

260

This data clearly continues to show a more marked use of the procedures nearing the conservation extreme of the continuum in both languages. Borrowing and fixed translation are again the most prominent procedures in both TTs. In a striking similarity to figures from Amanda’s Wedding and My Life on a Plate, the relative frequency of these two procedures is again higher for the Spanish language, with

55% borrowing and 21% fixed translation, compared to the Italian, 50% borrowing and 20% fixed translation. Literal translation follows as the third most popular procedure for this book in Spanish with a relative frequency of 11%. However, this procedure falls to fourth place in the Italian version, where generalization ranks third with a relative frequency of 10%, a significant difference if compared to the Italian relative frequency for literal translation which lies below 7%. Similarly to what happened in previous books, the difference with literal translation is also more marked between the languages, (71 instances in Spanish compared to 43 in Italian).

Notwithstanding the differences, these three procedures, borrowing, fixed translation and literal translation, cover almost 80% of all the translation instances in both TTs.

Regarding the other procedures in the conservation extreme of the continuum, gloss and explicitation are hardly used, with a relative frequency of just over 1% in

Spanish and even below the 1% mark in Italian. The relative frequency of adaptation is higher for both TTs, covering over 3% of all the translation instances in

Italian and 2% in Spanish. Of all the adaptations, nicknames again make for the majority of the examples in both languages, similarly to what happened in previous books:

261

(a) SL Spanish

Bee Renacuajo

Dusky Beauty Belleza Morena

Waspy Ratón

(b) SL Italian

Bee Apina

Dusky Beauty Bella Tenebrosa

Waspy Vespino

In the case of the procedures near the substitution extreme of the continuum, much more variation is present in the Italian TT than in the Spanish TT, and Italian shows consistent higher relative frequency than Spanish in all the procedures in this extreme.

Generalization is the most commonly used procedure of this group in both languages

– however the percentage in Italian is double that in Spanish. That is consistent with findings from Amanda’s Wedding, where again generalization showed a much higher frequency in the Italian TT.

Omission is used in 4% of all the translation instances in Italian but in less than 1% of the Spanish.

262

Cultural substitution shows a 2% relative frequency in Spanish and 3% in Italian.

Within this procedure changes in television programmes cover for the most of the examples:

(a) SL Spanish

EastEnders Vecinos

Casualty Urgencias

The Slobs Los vagos

(b) SL Italian

EastEnders Love Boat

Casualty I casi della vita

The Slobs Gli sciattoni

Finally, returning to the figures, explicitation, cultural replacement and insertion are barely used in either language, amounting to just over or just under 1% in all cases.

In summary, this data presents a stronger tendency to manipulate the names in the

Italian text than in the Spanish, as the graph below illustrates. Although the vast majority of the names are consistently translated with procedures from the conservation extreme of the continuum in both languages, the usage of procedures on the opposite extreme is more marked, with percentages of usage consistently higher, in the Italian text. These findings will be extremely significant in the discussion that follows in Chapter 10. This goes in line with data extracted from

263

Amanda’s Wedding, where more manipulation was present in the Italian than the

Spanish text. The key difference with book 2 seems to be the existence of an

external glossary in the Italian language– this feature seems to spread the other

names more evenly across the other procedures for this language. Data also seems

to reveal or suggest that the Spanish translator tended to substitute more in My Life

on a Plate than it is the norm for that language as indicated by the figures from

Amanda’s Wedding and Bridget Jones’s Diary.

Conservation Substitution

Figure 9.16: Procedure relative frequency against continuum – Bridget Jones’s Diary

264

9.2.6 Overall summary

Certain patterns of behaviour have started to emerge in this section. Borrowing, fixed translation and literal translation stand out as the most prominent procedures overall, covering almost 80% of all the translation instances in both languages. Of the other procedures, generalization, adaptation and gloss consistently show the highest relative frequency. All of these observations point perhaps to a combined tendency to conserve certain names with the mild substitution of some elements to make the text more clear or familiar for the target readers.

Notwithstanding the above comments, it is also true, however, that substantial differences between the individual patterns for each language can also be appreciated. Below, I will set out the tendencies by language, bringing out a more complex paradigm of behaviour around the translation of proper names.

9.2.6.1 Spanish

Looking at the data across all three Spanish TTs, borrowing shows a relative frequency of almost 60% completely dominating over all other procedures. By the same token, the majority of the procedures near the conservation extreme of the continuum (with the only exception of generalization) consistently show higher relative frequency than those at the opposite extreme. This, together with the low relative frequency for omission –just 2%-, highlight a strong tendency to maintain

265 the TT similar to the ST. Figure 9.17 below, shows the distribution of all the procedures for the three Spanish TTs:

2.37% 0.25% Spanish 1.00% 0.25% 2.69% 4.31% Borrowing 1.12% 2.87% Fixed Translation Transcription Literal Translation 10.74% Gloss (intra/Extra)

0.12% Explicitation 57.84% Adaptation 16.43% Generalization Cultural substitution Cultural replacement Omission

Figure 9.17: Relative frequency for procedures in Spanish TTs

A difference was noted, though, between the patterns for Amanda’s Wedding and

Bridget Jones’s Diary and those for My Life on a Plate. This book seems to show more variation in the data, and indeed, most of the occurrences of the procedures around the substitution extreme of the continuum take place mainly on this book.

The clearest example of this is omission, which is hardly used in Amanda’s Wedding or Bridget Jones’s Diary but has 5% relative frequency, 28 raw frequency, in My Life on a Plate.

There are various possible explanations for this, possibly the editor and publishing companies had an input on this. It could also be related to the author of the original

266 book, India Knight, a well-known journalist in the UK. This is a factor that might have perhaps inclined the translators to explore British culture more than a novel

Chick Lit writer would have done. It also stands out, for instance, the high number of extralingual glosses in Amanda’s Wedding, with a raw frequency of 21, where the translators choose to include a footnote to explain certain names to their target audience, for example they explain that Colditz fever refers to the Colditz palace used by the Nazis as a prison during World War II, or that Celebrity Squares is a popular British quiz show. In a similar way, the translator’s own preference could have played an important role in deciding the shifts in My Life on a Plate, which do not seem to follow the general trend for the Spanish language. During the data collection stages of this project, I corresponded with editors and translators of Chick

Lit books. On one occasion, one prominent Spanish translator of Chick Lit books, showing all good will, kindly wished me luck with my choice of subject ‘although I doubt you will find anything as proper names are rarely translated’. This clearly describes the general attitude in Spain towards proper names, and adheres to the general findings of this study.

The translator’s own preferences and choices are possibly a very important factor in all these changes as well as within all the trends identified in this analysis. This point will be stressed again in the Limitations section of the conclusions chapter (11.5), but the translators personal choices cannot be overlooked here. It is important to point out that perhaps with more data the picture could be different (see also 11.5), notwithstanding this point, the tendencies discerned in this study are consistent and very strongly backed by data across all TTs in BRIDGET.

267

In keeping with the format used to present the data so far, the graph below plots

the relative frequencies of the procedures against the continuum. This illustrates

very clearly the trend in the Spanish language to remain around the conservation

extreme:

Spanish 70.00% 57.84% 60.00%

50.00%

40.00%

30.00%

20.00% 16.43% 10.74% 10.00% 2.87% 4.31% 2.37% 0.12% 1.12% 1.56% 2.12% 0.25% 0.25% 0.00% Substitution Conservationn

Figure 9.18: Procedure relative frequency in Spanish TTs against continuum

9.2.6.2 Italian

Although borrowing shows undoubtedly the highest relative frequency of all the

procedures also for the Italian language, the percentage remains close to 50%,

spreading the other figures more evenly across the continuum. The general trend

exposed by the data seems to indicate a tendency in the Italian texts to make more

use of the procedures nearing the substitution extreme of the continuum than that

268 identified in the Spanish data. A change in referent potentially indicates an aim to make the text more easily understandable for the target readers – it can be presupposed that the new referent is considered by the translator as more valid, better known in the target language culture. The raw and relative frequency for cultural substitution and cultural replacement, consistently around 2 or 3% with an average of 16 tokens per book between them, illustrates this. These procedures were hardly identified in the Spanish TTs.

Figure 9.19 below shows the distribution of procedures in the Italian TTs:

4.00% Italian

0.13% 1.56% 1.44% Borrowing Fixed Translation 3.69% 7.32% Transcription 1.06% Literal Translation Gloss (intra/Extra) 5.63% Explicitation 50.59% 8.44% Adaptation Generalization 0.00% Cultural substitution 16.14% Cultural replacement Omission

Figure 9.19: Relative frequency for procedures in Italian TTs

Of the procedures nearing the substitution extreme of the continuum it is, however, generalization the one that shows the highest frequency – indeed this procedure ranks third overall, above literal translation, both in Amanda’s Wedding and Bridget

269

Jones’s Diary. The tendency differs in My Life on a Plate; as we saw before, the existence of the glossary makes the number of instances of gloss rise exponentially in this book. In some respects, the sheer existence of such glossary is an interesting finding in itself. It highlights the prominent role that proper names play in these novels, up to the point where translators and/or editors feel the need to include a description of these names (figure 9.20 shows an excerpt from this glossary). This glossary may also have been devised by editors without input from the translator, as some of the terms included in it receive a different treatment on the actual body of the text, especially those with more than one token. For example, the glossary includes a definition of Billy Goat Gruff. In the ST this name is used when Clara compares one of her stepsisters when she comes back from a shopping spree

(wearing new shoes and carrying a number of designer bags) to Billy Goat Gruff.

However, the Italian translator omits the reference to this character and instead,

Clara compares her stepsister to the murderer in the fable tales. The result is that there is no reference to Billy Goat Gruff in the Italian text but the term still appears in the glossary. This could indicate that the terms to be included in the glossary were selected without the translator’s help or before the translation was completed with no final review taking place at the end.

270

Picture 9.20: Excerpt from My Life on a Plate’s glossary in Italian TT.

Omission is another procedure around the substitution extreme of the continuum which presents similar relative frequency to literal translation both in Amanda’s

Wedding and Bridget Jones’s Diary. Interestingly, there is very low omission in My

Life on a Plate, its relative frequency barely reaches 1%, this is very different to the other books, where it nears on average 6%.

Overall, what ultimately emerges from the data extracted from Italian TTs is a clear tendency to use a wider range of the procedures in the continuum. Although the highest relative frequencies are recorded around the conservation extreme of the continuum, the procedures near the opposite extreme get more occurrences than they did in the Spanish TTs, showing a tendency to substitute and change more elements in the Italian TTs, as illustrated in figure 9.21 below:

271

Italian 60.00% 50.59% 50.00%

40.00%

30.00%

20.00% 16.14%

8.44% 7.32% 10.00% 5.63% 3.69% 4.00% 0.00% 1.06% 1.56% 1.44% 0.13% 0.00%

Conservation Substitution

Figure 9.21: Procedure relative frequency in Italian TTs against continuum

9.2.7 Conclusions

On the evidence supplied so far a pattern of the translation of names in Chick Lit is

starting to emerge, mainly presenting a mixture between the conservation of some

elements and the substitution of others. The next section explores the use of

procedures from a different perspective; looking into the different types of names

set in the taxonomy for proper names presented in Chapter 7, and trying to decipher

any possible relation between those names which remain unchanged or those that

are substituted and the different types of names described in the taxonomy. Is there

any relationship between name type and procedure used?

272

9.3 Procedures & Taxonomy

9.3.1 Introduction

All the names extracted from STs in BRIDGET and their translations in the TTs were classified according to the taxonomy for proper names presented in Chapter 7. This classification is now combined with the classification of translation procedures employed to translate the SL names (Chapter 8) with a view to determining possible patterns between name type and procedure usage.

Maths functions in Excel, such as SUMIF to count all the occurrences of a term when a certain condition is met, as well as relational tables in Access, with links between each table and term within the database, were used in order to combine both classifications. In this way it was possible to count the number of times a procedure is used within each individual category within each of the four facets in the taxonomy, for instance the number of times borrowing is used with ergonyms brand names in the ontological facet. A comparison between the figures obtained at category level with those obtained at procedure level (section 9.2 above) highlights any differences which might be due to name category.

In line with general findings presented at procedure level, in most cases the patterns uncovered by facet present a similar picture with higher frequency for the procedures around the conservation extreme of the continuum in almost all the categories. However, interesting patterns are present and very visible for certain categories, where the usage differs greatly from that of the general trend. These

273 fluctuations are specially marked with primarily denotative names in the semantic facet. Chrematonyms, ergonyms and pragnonyms in the ontological facet also show some degree of variation. This is consistent in both languages and across all three books.

The complete list of raw and relative frequency figures, percentages and graphs can be found at appendixes 2, 3, 4. The information, summarized below, is presented in general across all three books, by individual category in each facet with special notes about discrepancies in between the languages. Special attention is given to the categories where more variation occurs. The patterns concerning proper name types and translation procedure are also highlighted.

9.3.2 Ontological facet

9.3.2.1 Anthroponyms

Anthroponyms, personal names, present a very similar pattern to that of general procedure usage in both languages. Borrowing is the most prominent procedure across all three books in both languages, with percentages slightly higher than those found at procedure level. This is to the detriment of fixed translations, which fall slightly compared to the numbers found at procedure level.

From the qualitative exploration of the anthroponyms (details in section 9.4) it becomes apparent that personal names are the category more commonly used for

274 their connotations. Borrowing is the procedure used for the names of all the characters in the books. However, connotative personal names tend to be translated with other procedures. The data shows a high degree of usage of the procedure gloss (both adding information to the body of the text or as a footnote). Cultural replacement is also prominently used with these names as is adaptation. Section 9.4 will explore these patterns more in detail with the analysis of specific examples from the text.

9.3.2.2 Toponyms

In the case of toponyms (names of places) by contrast, fixed translation sees an increase in frequency consistently across all data for all books and languages. This is only to be expected, as names of places tend to have an established translation in each language and as a translation rule (Santoyo 1988), this form tends to be maintained (see 3.3 and 8.2.4), for example London/Londres/Londra or

América/Estados Unidos/America.

Within this category, the use of gloss in My Life on a Plate in Italian stands out, however, as explained in section 9.2.4, this relates to the glossary included at the end of the novel.

Also in the Italian language, it is worth mentioning the higher percentage of omissions found for this category in Bridget Jones’s Diary. These mainly correspond to entertainment or retail premises where the translator omits either the given name

275 or the reference all together, for example Harvey Nichols/dai vestiti or Warehouse/Ø.

This goes in line with findings highlighted in section 9.2.5, where the use of omission for the Italian language in this book was high overall.

All in all, from this data it can be said that after borrowing, fixed translation is the most commonly used procedure to translate names of places. This mainly applies to names of continents, countries, cities and physical features of the earth such as names of rivers, mountains, lakes and so on, where there is a translation already established in the target language. On the other hand, for names of other entities, more specific to the target culture such as retail or entertainment premises, more variation in the procedures used can be observed. For these more connotative entities, the data highlights the use of omission for book 2 in Italian, but the use of cultural substitution, gloss and explicitation is also more marked than in the data at procedure level.

9.3.2.3 Pragnonyms

The case of pragnonyms (names of events and phenomena) stands out for the extremely high percentage of fixed translations and literal translations. This can be interpreted two ways: firstly it has to be said that overall there are not very many pragnonyms in the corpus, only 48 in total. Such a relatively small number perhaps does not offer a representative sample. On the other hand, it seems relevant that a consistent picture occurs across all the names for events and phenomena. This can be related to the nature of pragnonyms, similarly to what was evident with

276 toponyms, which often have a standard translation which tends to be respected by the different translators. Examples include Christmas/Navidad/Natale, for instance.

The majority of the pragnonyms occur in Bridget Jones’s Diary, which is set around a calendar year, marking special celebrations like Christmas, Valentine’s Day or Easter, for example, all of which have fixed forms in the target languages.

Therefore, although the numbers are limited, the fact that the overwhelming majority of the cases were translated with the same procedures in both languages represents an important feature, consequently pointing to fixed translation and literal translation as the preferred procedures for this type of names.

9.3.2.4 Chrematonyms

The case of chrematonyms is perhaps more significant, as, although still low in number, there is a higher proportion of this type of name in the corpus, a total of

98. The first thing that immediately stands out with this category is the drop in borrowings, which fall well below the 50% mark that was consistently identified at procedure usage level. Most significantly, Bridget Jones’s Diary drops to around the

25% mark in both languages and My Life on a Plate plummets to 6% in Italian. The numbers spread across all other procedures more or less evenly in the Spanish language in Amanda’s Wedding and Bridget Jones’s Diary and show higher percentages around explicitation and generalization in My Life on a Plate.

277

For the Italian language, on the other hand, the numbers are higher around generalization and omission in Amanda’s Wedding and around literal translation and cultural replacement in Bridget Jones’s Diary.

Chrematonyms are names of politico-economic or commercial or cultural institutions, organisations, groupings or gatherings, so it is perhaps to be expected to see a wide range of procedures used with this category which incorporates such a wide range of entities. For example, in the case of some international, well known organisations, they will have their set names in the target language, and this form is respected by the translators, for instance The European Court of Human Rights Court/Tribunal

Europeo de Derechos Humanos/Corte Europea per I Diorite dell'Uomo; in the case of music groups the majority borrow the name in all languages, for example Spice

Girls, or the Human League; British organizations are adapted or translated literally, like RSPCA/sociedad protectora de animales, the samaritans/I samaritani. These variations in the types within the hypertype chrematonym account for the wide spread of procedures employed to translate them.

9.3.2.5 Ergonyms

Ergonyms, product names, make up over 27% of all the names extracted from the corpus, making them a more representative sample. The most striking feature in this category is the increase in the number of generalizations, which consistently show higher frequency than that recorded for this procedure overall. This happens in all books in both languages. This implies that most product names tend to be

278 standardized and turn into a bland element which cannot be attributed to any specific culture in particular. Examples include, for instance

Topics/chocolatinas/cioccolata, Hoover/la aspiradora/ l’aspirapolvere, King

Edwards/patatas/patate.

9.3.3 Scope Facet

This facet identifies whether the influence of the name falls only within its context in the novel or, on the contrary, whether it goes beyond the borders of the novel to have a wider meaning in the source culture or further repercussions outside the book. The pattern of translation for both categories in this facet is largely similar to that identified at procedure level.

9.3.3.1 External

Within external names, the high number of generalizations stands out in all books in the Spanish language and in Amanda’s Wedding and Bridget Jones’s Diary in the

Italian language. Consistent with previous findings, My Life on a Plate shows a higher number of glosses in the Italian language. These findings implies that for those names specific to the source culture the translators tend to either add some information in order to help guide the target readers to get closer to the referent and its connotations or they decide instead to use a more general term, no so easily attributable to the source culture.

279

9.3.3.2 Internal

For internal names the high number of adaptations seems relevant, a feature not so strongly present in any other category. The majority of internal names translated with an adaptation correspond to nicknames, and this seems to be another confirmed trend, as already exposed in section 9.2; translators seem to tend to adapt the nicknames using similar nicknames belonging to their target language culture: Sweet One/Bomboncito/Coccolona, Waspy/ Ratón/Vespino.

This is another important pattern highlighted in this exploration of name type and translation procedure usage: internal, connotative nicknames (anthroponyms) tend to be adapted to the target culture.

9.3.4 Sphere Facet

This facet reflects whether or not the referents of the names exist as entities in the real world. There are three categories in this facet: realis, irrealis and mythical.

However, there are only a handful of mythical names in the corpus (17). Although such small numbers could be considered to be subject to error it seems relevant to look at them more closely.

280

9.3.4.1 Mythical

The low number of occurrences of this name in the corpus (17) can possibly be linked with the category in itself rather than to a low representation of this type of name in BRIDGET, as mythical entities are scarce per se. Thus it seems pertinent to mention that the vast majority of these names were translated with a fixed translation, respecting the corresponding established name for these entities in the

TL, a relevant finding on its own. Examples from the corpus include, for instance, the Antichrist / L’Anticristo .

Within the other two categories, realis and irrealis, the patterns identified are very similar to the general patterns, with high volume of borrowings and literal translations in both categories across all books and languages.

9.3.4.2 Irrealis

Within the irrealis names in the Spanish language, the number of adaptations stands out – this again is linked to the treatment of nicknames as words of endearment, which tend to be adapted to similar forms used in the TL cultures, for instance

Bunny Pants/ Gran Dama/ Patatina ( already pointed out in 9.2.4 and 9.2.5).

In the Italian language the drop in generalization stands out in irrealis names, as does a slight drop in gloss in My Life on a Plate, compared to general trend in this book. This is perhaps to be expected as irrealis names belong to the actual book and

281 do not necessarily form part of the outside world and therefore there is no need to include them in a glossary explaining the terms belonging to source language culture.

9.3.4.3 Realis

For realis names, there is a slight rise in the use of generalization in both languages.

And the number of glosses in My Life on a Plate in Italian grows too, more in line with both general trend identified and with the nature of the category and the facet.

In summary, for the sphere facet the patterns highlighted are as follows:

Fixed translation is the most commonly used procedure for mythical names.

Within the internal names, the data shows that nicknames tend to be adapted to the target culture. For the rest of internal names, borrowing and literal translation are the most commonly used procedures, with higher usage than the overall at procedure level. As these names belong to the particular book and are linked to the context and plot of the novel, there is no perceived need to add any gloss or extra information or to adapt or substitute the name. On the other hand, realis names see a higher percentage of usage of the procedure generalization.

9.3.5 Semantic facet

This is the facet where more variation can be found compared to the general trend identified at procedure level. There are fluctuations in all three categories in the facet: primarily denotative, intermediate, primarily connotative. It is in this last

282 category where the differences are more marked. Chapter 3 presented the opposed views taken by scholars interested in the translation of proper names along the years. These scholars have argued about the possibility or impossibility of the translation of proper names. The consensus for all of them was found, though, around connotative names, always considered as translatable items. The figures collected for this facet in general and specifically for the connotative names show that this position is still active nowadays, with more changes occurring with these types of names

9.3.5.1 Primarily denotative

Primarily denotative names show higher percentages of borrowing and fixed translations in all three books in both languages. Literal translation numbers, on the other hand, are lower. The other procedures are more or less consistent with general trends at procedure usage level.

9.3.5.2 Intermediate

Intermediate names also show similarities with the general trends at procedure level. A surge in the number of generalizations stands out in both languages. This reflects a tendency to take those names with certain degree of connotations closer to the target language, facilitating the task of interpretation for the reader.

9.3.5.3 Primarily connotative

283

This tendency to manipulate the names is clearly more marked with primarily connotative names. The main function of these names is to point to the characteristics associated with the referent of the name. The idea is that recognition of the referent behind the proper name triggers an association which then helps to guide the reaction of the reader. The numbers show more variation in the procedures used to translate this type of name than with any other category in the taxonomy.

The number of borrowings drops considerably, to well below the 50% mark where it usually lies. On the other hand, the procedures around the substitution extreme of the continuum see a higher degree of use. Literal translation, nearer the conservation extreme, is also used more.

This shows, again, a tendency to guide the target readers with those names which are used for their connotations, expecting to create a reaction in the source language readers – translators use different tools in order to help recreate these reactions in their target audiences.

In summary for this facet, primarily denotative names are mostly borrowed or literally translated. Intermediate names tend more towards generalization whereas cultural substitution and replacement, adaptation and omission all see a surge in usage with these names compared to the figures at procedure level.

284

9.3.6 Conclusions

The concrete patterns that this data seems to be pointing towards are as follows:

 Anthroponyms – tend to be used for their connotations more than any other

type of name.

Names of characters are borrowed.

To translate more connotative personal names, specific to the target culture,

gloss, adaptation and cultural replacement tend to be the procedures used.

 Toponyms

Names of geographical features use the fixed translation established in the

target culture.

More culturally specific names such as retail or entertainment premises tend

towards cultural substitution, gloss or omission.

 Chrematonyms do not seem to be borrowed as much as other type of names.

In contrast, a wide variety of procedures is used with this type of name.

Adaptation and cultural replacement stand out.

 Ergonyms tend towards generalization or fixed translation.

 Pragnonyms tend towards literal or fixed translations

 Scope and sphere facets see similar patterns:

- External/Realis names tend towards generalization

- Internal/irrealis names (primarily covering for the nicknames given to the

characters in the books) tend towards adaptation.

 In the semantic facet connotative names consistently tend to be modified with

procedures around the substitution extreme of the continuum.

285

The analysis of procedure usage per category in the taxonomy has shown more manipulation taking place with certain types of names such as anthroponyms, chrematonyms or ergonyms but above all and more markedly with connotative names. This is a crucial point and seems central to uncovering the decision-making processes underpinning the selection of procedures. The tendency identified for primarily connotative names can only really be understood with examples in the context in which they are used in the novel. As these names are so rooted in the source language culture and so embedded in what is happening in the novel, they go hand in hand with the cultural and linguistic contexts in which they occur. In the next section, I present a selection of primarily connotative names, and explore their contexts and the procedures used for their translation in both languages. The purpose of this exercise is to try and understand the SL names in their context and the full extent of their connotations for the SL readers. It also serves as a way to observe the possible reasons behind the selection of procedures employed by the different translators and opens the way to exploring up to what extent they manage to evoke similar feelings in the TL audience.

9.4 Names in context – qualitative analysis

9.4.1 Introduction

This section weaves in material from all books in the corpus, exploring the different ways in which connotative names are used in the SL texts as well as the different methods employed for their translations. The names selected for commentary have

286 been chosen because of their particular usefulness in illustrating the way in which proper names are employed primarily for their connotations. The five hypertypes in the ontological facet (anthroponym, toponym, ergonym, chrematonym and pragnonym) are taken as the axis around which these primarily connotative names are arranged. This approach contributes to shedding new light on the study of the relationship between name category and translation procedure, and the sections are directly linked to the patterns highlighted in 9.3 above. At the same time a more qualitative analysis of the names in context transpires as key to understanding up to what point the cultural background and sense of place evoked for SL readers by a proper name can be translated into different target languages with different traditions, perceptions and values.

The system used to describe each name is consistent for all the examples, introducing the terms and procedures in a table with five columns like this one:

SL (a) Procedure (b) Procedure Book Title Spanish Used in SP TT Italian Used in IT TT Book Title Book title

The first column to the left contains the SL name, followed by example (a) in

Spanish and the procedure employed with it, then is example (b) in Italian and the procedure used. The title of the book where the name originates is given in the headings raw, under each corresponding language. For ease of reference, as well as these tables with the individual lexical units the concordance line for each term is also provided. The commentary which follows describes the context and the elements at play for each name and looks into each individual translation.

287

9.4.2 Anthroponyms

The tendencies with anthroponyms that the data analysed in section 9.3 showed were twofold, on the one hand the use of borrowing to translate the names of the characters was prominent, whereas, on the other hand a tendency towards a use of other procedures at the other extreme of the continuum with connotative personal names was also clearly visible. The examples collected in this section in particular highlight this pattern as the vast majority of the examples have been adapted, modified or replaced in an attempt to offering more information to the target language reader.

The anthroponyms within the primarily connotative names tend to be used in a variety of forms such as nicknames constructed by puns or names of entertainment or cultural figures used as comparatives or premodifiers to describe some feature of one of the characters, perhaps a trait, or a mood, a way of being or a particular feeling or style. Knowledge of the referent as well as of the connotations associated with it is paramount in order to fully understand its meaning in the context. The names below collect some of these different uses of personal names, and illustrate some of the different approaches, explored in section 9.3, taken by translators in their attempt to transfer both referent and sense of the name.

288

Example 1 – Pun

SL (a) (b) Amanda’s Spanish Italian Procedure Procedure Wedding La boda de Non sposate Amanda quella donna! ¡No llevo Knickerless Explicitation Ø Omission bragas!

Concordance Line 1

English Spanish Italian Si Fran no hubiese calificado de gilipollas a todos los tíos que habíamos mencionado en al última hora y media, quizá Had Fran not said wanker le habría hecho caso y me Se Fran non avesse dato del about every bloke we’d habría ahorrado un trauma. coglione a tuttil gui uomini che mentioned for the last hour Sin embargo, me puse a avevo nominato nell’ultima ora e and a half I might have listened hacerle señas trazando mezzo, forse le avrei dato ascolto to her and saved myself some enormes círculos en el aire. e mi sarei risparmiata qualche trauma. Instead, I waved at him <<¡No llevo bragas!>> Y me trauma. Invece mi sbracciai nella in huge circular motions. deshice en una risa nerviosa. direzione di lui, scoppiando in “Kickerless!” And I disolved in Se acercó volando y me dio un imbarazznti risantine. Lui si giggles. He flew over and gave beso grande. Oh, sí que precipito verso di me e mi dette me a big Kiss. Oh, we must debíamos de ser viejos un baccinoe. Dovevamo essere have been old friends, then. amigos, por aquel entonces. grandi amici, allora.

The first example is a play on words, with the name of one of the characters as the bases for the joke. The action takes place at the beginning of the novel. The main character, Melanie, and her best friend, Fran, have gone partying. They are drinking at a bar when they spot Nicholas, Melanie’s occasional lover, in the crowd. They decide to attract his attention in order to get him to buy them some drinks, so

Melanie shouts ‘Knickerless’ smiling at him.

This homophone blends the similar pronunciation of the man’s first name with a compound to mean that Melanie is wearing no underwear. This form presents special difficulty both for Spanish and Italian translators, not only because the

289 original relies upon formal phonic similarities between the units, but most importantly because their TLs do not tend to use agglutination to form new words often.

Thus, the Italian translator chooses to omit the name. The TT recollects all the action, with the girls seeing Nicholas and approaching him but the pun is omitted;

Melanie simply goes to him with a big smile instead of calling out to him. All the humour is consequently lost, and there is no attempt to compensate for this loss.

The Spanish translation, on the other hand, takes a very different form, with the translator opting to explicate the sense of the pun. Melanie literally shouts I am not wearing any knickers - the whole meaning is spelt out, by means of an explicitation, without introducing a play on words, by the same token, the reference to the character’s first name is lost and the whole sentence seems to relate solely to the jokey and current playful mood of the character, leaving the specific pun out.

Example 2 – Fictional character catch phrase/special trait

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! (Una dura Paddington mirada de) Uno sguardo Bear (hard Literal translation Explicitation Osito da iceberg stare) Paddington

Concordance line 2

English Spanish Italian Fran gave him a Paddington Fran le lanzó una dura mirada Fran gli rivolse uno sguardo da Bear hard stare. de osito Paddington. iceberg.

290

Following on from the previous scene, Nicholas and Melanie spend the night together at her house. The morning after, Fran comes to visit early and the girls do not know what to do to eject him. An unforeseen turn of events follows when the postman delivers a postcard from Melanie’s long-lost boyfriend, who had unexpectedly emigrated to America the previous year, without giving her any notice.

He announces in his postcard that he is coming back to England. Melanie is in a state of shock, while Fran tries to take control of the situation. In the middle of all this, Nicholas is intruding with questions and does not realise he is not welcome anymore, until Fran very clearly with a ‘Paddington Bear hard stare’ sends him to buy some chocolates.

The name of the famous bear is used here together with one of his best known traits – the hard stares that he offers when he is upset, or unhappy- as a way to exemplify the hard approach that Fran has to take on Nicholas, who does not get any of Melanie’s subtle attempts to see him off.

The Spanish TT shows a literal translation of the original. However, this attempt is unsuccessful at recreating the effect of the original. Dura mirada [hard stare] and osito [teddy bear] do not work together in the Spanish language or culture, where this trait of Paddington bear is not so well known and an osito always represents a cute cuddly toy. Without the clear reference to the specific trait of this one particular bear this phrase loses its meaning, and the Spanish audience are left with a phrase combining two opposing elements.

291

The Italian TT, on the contrary, literally explicitates the meaning of the original, omitting the reference to the bear and simply describing the stare as a hard, cold one. Here, the British cultural background of the original is lost, but the connotations are maintained and the phrase recreates instance the way in which Fran is treating

Nicholas.

Example 3 – Entertainment figure as comparative

SL Spanish Italian Bridget Jones’s El diario de Procedure Il diario di Procedure Diary Bridget Jones Bridget Jones Jeremy La Adaptation Una pacata Adaptation Paxman femminista

Concordance line 3

English Spanish Italian 'What about you, this week? —¿Y tú qué, esta semana? "E tu allora, questa settimana? Prima First you completely ignore Primero me ignoras por mi ignori completamente come una me like some Hitler Youth completo como una solterona vergine di ferro delle gioventù ice-maiden, then you turn de hielo de las juventudes hitleriana, poi ti trasforma in una into an irresistible sex Hitlerianas, luego te conviertes irresistibile gattina tutta sesso, kitten, looking at me over en una irresistible y sexy gatita, guardandomi da sopra il computer the computer with not so mirándome por encima del con due occhi che parevano dire non much "come-to-bed" as just ordenador con ojos no tanto de tanto 'vieni a letto con me' quarto "come" eyes, and now «ven-a-la-cama» como de sólo 'viera e basta'. E adesso tutto a un suddenly you're Jeremy «ven», y ahora, de repente, tracto ti metía a fare la femminista." Paxman.' eres una pacata.

This example belongs to Bridget Jones’s Diary and takes the form of a comparison with an external realis entertainment figure, using characteristics specific to this figure in order to describe the behaviour of one of the characters.

292

In the novel, Bridget has adopted the strategy to ignore Daniel in order to arouse his attention. Daniel finds this sudden change in Bridget’s attitude disconcerting and confronts her about it: ‘one day you are all over me like a…. and today you are acting like Jeremy Paxman’. The author has chosen to use the connotations around the figure of this journalist, known for his direct approach and straight and formal style, to illustrate Bridget’s behaviour instead of using more explicit adjectives.

Knowledge of these connotations associated with Mr Paxman’s personality is, of course, essential.

Both translators choose to adapt this name to their target language. In both TTs the referent is lost, with no clear entity behind the new word; they rather use a general noun to describe more specifically, as opposed to subtly, Daniel’s feelings. The

Spanish term pacata is an unusual word, very rarely used in day-to-day language. It describes someone who is timid, too scrupulous or modest. In a way, Bridget is behaving in a timid or prudish way, ignoring all of Daniel advances and remaining professional and distant, but the whole extent of the connotations implied with the comparison to Jeremy Paxman is lost in this translation. By the same token, they are also lost in the Italian version, where the adaptation also sees the referent disappear in favour of a general noun, femminista, describing Bridget as a feminist. Again, to some extent this word serves to describe Bridget’s attitude, ignoring Daniel, remaining cold, but the connotations implied are very different to those evoked for

British readers by Jeremy Paxman.

293

Example 4 – Entertainment figure as comparative

SL Spanish Italian My Life on a Mi vida sobre un Procedure La mia vita su Procedure Plate plato un piatto Doris Day Doris Day Borrowing Doris Day Borrowing The Doris El estado de La Doris in Adaptation Adaptation Streak ánimo Doris me

Concordance Line 4

English Spanish Italian None of these thoughts are Ninguno de estos pensamientos Nessuna di queste idee giova alla good for my Doris Day streak, son buenos para mi estado de Doris Day che c'è in me e, che, which has come skipping and ánimo Doris Day, que se ha ido tutta saltelli e percalle, in questi ginghaming its way to the manifestando y fortaleciendo ultimi giorni ha p reso il fore in recent days. I hate my durante los últimos días. Odio mi sopravvento. Io odio questi streak, but I can't help having estado de ánimo, pero no puedo momenti, ma non posso it, nor it me, and I've given up esquivarlo, ni él a mí, y ya he impedirmi di averne (né, loro, di on trying to shake it off. desistido de sacudírmelo de avere me) e ormai ho smesso di encima. cercare di farmeli passare.

This example is based in a particular trait of a famous actress, using it to describe the character’s state of mind at the present moment.

We are now in My Life on a Plate, and at a certain point in the novel the character,

Clare, compares herself with the actress Doris Day. What Clara defines as ‘my Doris

Steak’ sees her loving all natural things, trees, plants, rocks, and being ethereal and unmaterialistic. This relates to the actress’s love for animals and her altruistic work to protect them and their environment.

Both TTs maintain the reference to the actress, presumptively, on the basis that she is a well enough known celebrity, familiar for both TTs readers. The second

294 example, the Doris Streak, shows a move from both translators to adapt the sentence to the forms and rules of their TLs, transforming the phrase to a valid one in the grammar and culture or their TLs, but still maintaining the referent of the original.

Example 5 – Entertainment figure as comparative

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! una delle Cultural Mystic Meg Una adivina Generalization Charlie's replacement Angels

Concordance Line 5

English Spanish Italian Chi credevo di essere, una delle Who did I think I was, Mystic ¿Quién me creía? ¿Una Charlie's Angels? Perché diavolo Meg? What had I done this for? adivina? ¿Por qué había hecho l'avevo fatto? Cominciai Silently, I began to cry. This eso? En silencio, me puse a silenziosamente a piangere. lovely, gorgeous bloke was still llorar. Este tío guapísimo y Quest'uomo bello e adorabile si going to get married, and he encantador iba a casarse de sarebbe sposato lo stesso, e in was going to hate me into the todos modos, y de paso iba a più mi avrebbe odiato per bargain. odiarme. sempre.

Again, this example relies on the readers’ knowledge of the connotations associated with an entertainment figure. Back with Amanda’s Wedding, and half way through the novel the characters are desperately trying to stop the wedding between Fraser and Amanda from taking place. In an attempt to make the groom see the real nature of the bride, Melanie recorded a conversation with her, where she admits she is mainly interested in Fraser’s money and the Laird title he has recently inherited from his grandfather. At this point in the novel, Mel has played the tape for Fraser

295 and he reacts angrily. She feels really bad about what she has done, and thinks to herself ‘who did I think I was, Mystic Meg?’

The author uses this famous astrologer and psychic to describe the way Melanie is feeling and how she sees herself at that point. Knowledge of the referent is paramount in order to understand the comparison and what the author is trying to illustrate.

Both Spanish and Italian TTs undergo a change in referent. In the Spanish version the proper noun is replaced with a common noun, and the specific referent is lost.

Una adivida [fortune-teller] encompasses the profession for which Mystic Meg is most famous for, but the other connotations around her are void in this translation.

The Italian translator, on the other hand, chooses to replace Mystic Meg, with una delle Charlie’s Angels [one of the Charlie’s Angels] still with a foreign feeling, but from a very different background. The Italian version, takes as its referent one of the private investigators in the famous American crime fiction series and film. The new name still maintains a foreign flavour, albeit not a British one, at the same time that it evokes certain connotations, of secret investigations and plots. The new referent presumably is more easily accessible for the Italian audience, and its connotations more widely understood.

296

Example 6 – Nickname constructed by pun

SL Spanish Italian Procedure Procedure Bridget Jones’s El diario de Il diario di Diary Bridget Jones Bridget Jones Pukey Ø Omission Succhia Adaptation

Pukey Pukey Borrowing Succhia Adaptation

Concordance Line 6

English Spanish Italian —¿Suki? ¿Cómo puede andar por el mundo con este "Suki? Succhia, semmai", l'ho 'Suki? Pukey, more like,' I said, nombre? —dije, convencida interrotto, pensando che stesse thinking he was about to say, de que él estaba a punto de per propinarmi la scoria del 'are brother and sister,' decir «somos hermanos», fratello e sorella, dei cugini, degli 'cousins,' 'bitter enemies,' or «primos», «enemigos a incompatibili o dell'acqua 'history.' Instead he looked muerte» o «historia». En passata. Invece I'ho visto rather cross. cambio, pareció enfadarse. piuttosto seccato."

He oído a Daniel, quedando I heard Daniel on the phone por teléfono con Pukey para "L'ho sentito parlare al telefono arranging to meet Pukey esta noche, y ha dicho con una con Succhia per combinare di tonight and he said in a flopsy- vocecita dulzona: «No vedersi stasera, e con una voce bunny voice, 'Not too bad... so demasiado mal... por ahora»,y tutta melliflua luì le diceva: far,' and I knew he was talking he sabido que estaba ""Non come temevo... per about my reaction, as if I were hablando de mi reacción, adesso"". Ho subito capito che Sara bloody Keays or someone. como si yo fuese la jodida Sara alludeva ella mia reazione, Am seriously considering face- Keays o alguien por el estilo. nemmeno fossi chissà chi. Sto lift. Estoy co seriamente pens

This example, extracted from Bridget Jones’s Diary, presents another wordplay, again relying in the similarities between certain words in the SL. Pukey is the form which Bridget has chosen to refer to Sukey, Daniel’s latest girlfriend – whom, needless to say, she dislikes. As just mentioned, this pun is based in the similarities, both graphic and phonetic, of the two words in the SL, playing with the real name

Sukey and the slang verb to puke – to vomit. Combined together they form Pukey,

297 to colligate the way this new girlfriend makes Bridget feel, with a clear allusion to the real name of the character.

There are two tokens of this name in the corpus – both are translated using the same cultural substitution in Italian. The translator uses the word succhia [dirty] in an attempt to recreate the original pun; playing with a word with a similar pronunciation to the original name, Sukey. The connotations differ to those evoked in the ST but the new name still evokes the negative feelings that Bridget feels towards Sukey. The Spanish translator omits the first reference to Pukey. When

Bridget is firstly introduced to Sukey in the original, she mentally starts calling her

Pukey; this does not happen in the Spanish translation, where the whole reference is omitted from the text. The next time that Bridget has an encounter with Sukey, she again uses Pukey instead of her real name. The Spanish translator uses a borrowing of Pukey this time. Both fail to recreate the connotations and in fact, the reference to Pukey in the second occurrence of the name comes out of the blue in the Spanish

TT, as the original reference, where the pun was firstly introduced, is not there as a guide. Both Sukey and Pukey are used in the Spanish text indiscriminately to refer to the same character without any explanation or reason for alternating them. It can be said that on this occasion the TL readers are left with two different names without a valid referent behind them with the consequent loss of cohesion in the text.

298

Example 7 – Pun

SL Spanish Italian My Life on a Mi vida sobre un Procedure La mia vita su Procedure Plate plato un piatto Pobre Nomes * Povera 1. Juego de Gnomi** palabras Altro gioco di intraducible: parole basato Nomes significa sull'omofonia di Poor Nomes Gloss Gloss nomos. (N. del Nomes, T.) nomignolo del personaggio, con gnomes, gnomi. [N.d.T.] Literal Rich Elves Ricos elfos Literal translation Ricchi Elfi translation

Concordance Line 7

English Spanish Italian

Hablando de mujeres débiles, Parlando di smutandamenti Speaking of feeble pant- ya es hora de que quede a facili, è arrivato il momento di droppers, it's about time I had comer con Naomi, la cual, pranzare con Naomi che, grazie a lunch with Naomi, who, thanks gracias al pichaloca de quel (letterale) testa di cazzo che to the literally dickheaded Richard, será conocida a partir è Richard, d'ora in poi sarà Richard, will now forever be de ahora por sus amigas como conosciuta presso gli amici come known as Poor Nomes by, her la Pobre Nomes * (en la Povera Gnomi** friends (as opposed to Rich contraposición a los ricos (presumibilmente in opposizione Elves, presumably). Elfos, supongo'). ai Ricchi Elfi)

This is another example of a nickname and a pun. This particular example has been extracted from My Life on a Plate. Naomi is married to a colleague of Clara’s husband. Clara is not very keen on her and sometimes uses the nickname Nomes to refer to her. Their children go to the same school and Naomi is always the perfect mother, never late for school, always perfectly well dressed and made up. Clara knows, however, that Naomi’s husband has been having an affair with his secretary.

Upon receiving the news of another of their acquaintances infidelity she wonders if everybody nowadays is being unfaithful. Her thoughts then take her back to Naomi,

299 who ‘thanks to the literally dickheaded Richard, will now forever be known as Poor

Nomes by her friends (as opposed to Rich Elves, presumably)’.

This is again a multiple pun (dickheaded Richard plays a part too) very difficult to recreate in the TLs. It is based on the homophones Nomes and gnomes, as well as on the homonymic of Rich, the husband and the adjective, both relying on similarities based on the semantic of the SL. Both translators opt for literal translations with a paratextual gloss, as a footnote, attempting to explain the pun.

This serves the purpose to describe the joke, but the reading flow on the main body of the text is broken and the playfulness of the original lost.

Example 8 – Nicknames constructed by pun

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! James La zorra Stronz Cultural Bitch Cassidy Literal translation Cassidy (agente replacement segreto) Bitchaham Zorraham Lyndon Cultural Literal translation Lincoln Lincoln Stronzon replacement

Concordance Line 8

English Spanish Italian -Si fuera vaquera; sería la zorra 'If she was a cowboy, she'd be Cassidy. -Fran; es una idea Se fosse un agente segreto, Bitch Cassidy.' `Fran, that's a bastante sutil. Pero Yo creo que sarebbe James Stronz. Fran, è pretty subtle idea. But I think Angus Seguramente ya se lo Ha un'idea arguta. Ma penso che Angus might have already tried it, dicho; de una manera u otra. Angus ci abbia già provato . somehow.' Angus half tilted his Angus asintió ligeramente con Angus fece un cenno con la head. la cabeza. testa.

300

'Look, why don't you just tell him. -Mira, ¿por qué no se lo Senti, perché non glielo dici? Go up and say, Fraser, don't cuentas? Ve y dile: Fraser, no te Vai da lui e digli: 'Fraser, non marry her, she's a bitch. She's a cases con ella, es una zorra. Es sposarla, è una stronza. È una skinny-rumped, dyed-haired bitch una culo flaco peliteñida zorra, stronza coi capelli tinti. È una bitch bitch. She's a complete and zorra, zorra. Es una completa y completa e purissima stronza. utter utter utter bitch. If she was total, total, total zorra. Si fuera Se fosse un presidente a president, she'd be Bitchaham presidente, se llamaría americano, sarebbe Lyndon Lincoln.' Zorraham Lincoln. Stronzon'.

These nicknames play once more with language forms. The examples are from

Amanda’s Wedding and again, the main characters are plotting to stop the wedding.

This time they are trying to make Fraser understand the kind of person Amanda truly is. They tell him that ‘if she were a cowboy she would be Bitch Cassidy’ and ‘if she were a president she would be ‘Bitchaham Lincoln’.

Here, the SL names involve the creation of new names by compounding referents to real entities –either the cowboy Butch Cassidy or the American president Abraham

Lincoln- with adjectives to describe Amanda’s true character based around the adjective bitch. The referents are universal enough, and presumably clear enough for the target audience, and the new compounds work to describe , perhaps in a more explicit way than in the ST, what they truly think of Amanda as much in the

TLs as they did in the SL.

The Spanish translator maintains the same referents, choosing to literally translate the compounds, playing with the word zorra [bitch] and the same surnames found in the original. The referents remain unchanged, and as considered universal enough, the connotations too should prevail for the TL audience. At the same time, the

301 choice of zorra brings the compound closer to the Spanish readers, both parts working together to evoke similar feelings in the TL readers.

The Italian translator again chooses to replace both referents. In the first example the cowboy is replaced by a reference to the secret agent James Bond. In the second example an American president is still used, but this time Abraham Lincoln is replaced by Lyndon Johnson.

Both work well with the Italian translation of bitch, stronza, to form a similar compound to that found in the original, but one valid in the Italian language. Despite the change in referent, both compounds manage to evoke similar connotations and the extent of Melanie’s and the other characters’ feeling for Amanda can be presupposed to be fully understood in both translations.

9.4.3 Toponyms

In the case of toponyms, the data in section 9.3 highlighted a marked use of fixed translation with the names of geographical features as well as with the names of countries, regions, continents and so on. More variety was identified with the other types of toponyms and those more specific to the source language culture; the examples collected in this section are a representation of these. There are 31 examples of primarily connotative toponyms in the corpus. The selection below shows the names in context at the same time that it explores the translations and the choice of procedures.

302

Example 9 Toponym as comparative

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! El internado Malory Cultural de Malory Gloss Eurodisney Towers replacement Towers

Concordance Line 9

English Spanish Italian Deja de ser tan amable Stop being so nice to me! What conmigo. ¿Qué era esto, el Piantala di essere gentile! was this, first day at Malory primer día en el internado de Cos'era, l'inaugurazione di Towers? Malory Towers? Eurodisney?

In this example from Amanda’s Wedding, Melanie has been demoted at work and relocated to a new office area within a new team. She is devastated about this change and dreading the move, but everyone is very friendly, in fact, too friendly and she feels as if she has arrived to ‘Malory Towers’. This name refers to the novels by Enid Blyton featuring the Malory Towers boarding school, where pupils learnt good values and achieved a great amount academically. The environment described is one of peace and a good atmosphere, welcoming of new students each year.

The Spanish version includes an intralingual gloss, adding the word internado

[boarding school] to make the referent more specific. This works well within the novel, without interrupting the reading process. However, the choice to maintain the

English name of the school, Malory Towers, instead of changing it for the name that the school receives in the Spanish translations of the original books, Torres de

Malory, might make it difficult for the TL readers to connect the name with its referent, and its connotations. This brings about a loss of intertextual cohesion 303 through the breaking or stretching of the networks of naming that might confuse the readers, on the one hand familiar with the referent but on the other presented with a foreign name perhaps unfamiliar to them.

The Italian version is very different, making use of a cultural replacement. Malory

Towers is replaced with the opening celebrations at Eurodisney. It is to be expected that for the Italian translator this would make a more easily recognisable referent for the target audience. To a certain extent, the new referent still portrays similar connotations to those of friendship and good will evoked in the original, while undeniably bringing new ones with it.

Example 10 Descriptive toponyms

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! Kennington Kennington Borrowing Kennington Borrowing Kensington Kensington Borrowing Kensington Borrowing SL Spanish Italian My Life on a Mi vida sobre un Procedure La mia vita su Procedure Plate plato un piatto Cotswolds Cotswolds Borrowing Cotwolds Gloss Crouch End Crouch End Borrowing Crouch End Borrowing

304

Concordance Line 10

English Spanish Italian She snorted with laughter. 'I'm Soltó una risotada. -¡Estoy more used to Kensington than más acostumbrada a Kennington!' she said Kensington que a Kennington! Lei rise. «Sono più abituata a hilariously. -dijo riéndose. Kensington che a Kennington! »

Su casa, por supuesto, es más Her home, of course, is more Cotswolds que Crouch End. Casa sua, naturalmente, è più in Cotswolds than Crouch End. Hay colchas antiguas y stile Cotswolds che C ro uch End. There are old quilts and bits of centones que cubren el Ci sono vecchie trapunte e avanzi patchwork blanket draped over respaldo de sus sofás, di copriletti patchwork the back of her cat-scratched atormentados por las uñas de drappeggiati sugli schienali dei sofas, wild flowers in lumpen los gatos, flores silvestres en divani graffi ati dai gatti, fiori pottery jugs, vaguely jarrones de cerámica selvatici in umili brocche di Bloomsburyish paintings on the panzudos, cuadros con un leve coccio e quadri vagamente walls, which are painte aire de Bloomsbury bloom

These examples, extracted from two different novels, illustrate one of the most typical uses of primarily connotative proper names, when the names are chosen mainly for their connotations, and the audience is expected to be familiar with these connotations and to interpret the message accordingly. In many cases, these books are as successful as their author is creative and able to identify referents universal enough in order to reach as wide an audience as possible. It is these connotative names which represent a particular problem for the translator, as authors no doubt consider their target audiences but not necessarily an international audience and some of the names, so rooted in the SL culture, are not easily understood outside the country of origin.

Both set of examples have been constructed in a similar way. Two very distinctive and different parts of the UK are being set against each other in order to make their differences stand out. The examples from Amanda’s Wedding, feature two London areas, the affluent Kensington set against the more working-class Kennington,

305 where Melanie lives. She has organised a dinner party and has been obliged to invite

Amanda and some of her middle-class friends. Mookie, one of these friends, arrives half an hour late and she excuses herself explaining that she got lost, as she is

‘more Kensington than Kennington’. This joke highlights the posh background of the girl at the same time that points to Amanda’s lower class in a scornful way. Of course, you have to be familiar with the characteristics and very different environments in these two parts of London to be able to understand the full extent of the sentence. It might be the case that these connotations are very specifically targeting Londoners as not every British citizen necessarily has to be familiar with the different backgrounds. This aspect adds more difficulty to the translation of these toponyms. Both translators, Spanish and Italian, opt to borrow the term, without attempting to explicate the meaning or to add any type of explanation or additional information. It is possible that they did not feel the need, or that they did not consider the action an important part of the core of the novel, and they were willing to simply introduce foreign, unfamiliar elements for their readers.

In the second set of examples, from My Life on a Plate, Clara is thinking about

Naomi whose house is ‘of course, more Cotswolds than Crouch End’. Here, again, two very different parts of the UK are taken as opposing entities to point out the type of house, and by assimilation also the type of person that Naomi is. A house in the Cotswolds brings to the imagination of the SL reader a little (or not so little rather) stone cottage, decorated with pretty flowers similar to what you often find depicted in a box of chocolate. The book once again is playing on stereotypes,

306 triggered implicitly in the ST, where they appeal to the inside knowledge of their readership to be fully understood.

Both translators again use borrowing as the procedure for these toponyms. It is interesting to point out, however, that the Italian version includes a description of the Cotswolds in its glossary of British terms at the end of the book. No entry, however, for Crouch End.

Example 11 Dialect variety

SL (a) Procedure (b) Procedure My Life on a Spanish Italian Plate Mi vida sobre un La mia vita su plato un piatto Landon Ø Omission Ø Omission

Concordance Line 11

English Spanish Italian Es fantástico estar en Londres -está diciendo Dunphy-. È grandioso essere a Londra, sta 'Great to be in Landon,' is Amedrentador... Un gran dicendo Dunphy. Imponente... saying. 'Intimidating... big reto... Me hospedo con unos una grande scena... stare con i venue... staying with friends.' amigos. miei amici.

This is a different example which plays with accents to mark differences in the pronunciation of the word London. In this particular case, from the novel My Life on a Plate, Landon is trying to evoke the Irish accent. The Irish modern dancer Sam

Dunphy explains that he is ‘very happy to be in Landon’ at the opening of his interview with Clara.

307

Again both translators take the same procedure, in this case omission. The dialect variety is omitted from both TTs, and the artist is simply happy to be in Londres in the Spanish TT or Londra in the Italian TT. There is no mark of the different pronunciation or accent of the man, no specific remark to his Irish background. This is lost in both translations.

9.4.4 Chrematonyms

There are 34 primarily connotative chrematonyms in BRIDGET. As identified in

9.3.2.4, these are translated with a wide range of procedures, from both extremes of the continuum. Borrowing is only used with 7 names, this figure is equal for both

TTs. The examples below represent some of the uses of connotative chrematonyms in context and explore the wide range of procedures employed for their translation.

Example 12 British Charity

(a) (b) SL Spanish Italian Amanda’s Procedure Procedure La boda de Non sposate Wedding Amanda quella donna! l'Associazione The El teléfono de Cultural dei Samaritani Gloss Samaritans la esperanza substitution

308

Concordance Line 12

English Spanish Italian

¡Agh! No aguantaba más. Iba a Aargh! Non c'era niente da fare. Arrgh! This was it. I was going tener que acabar llamando al Avrei dovuto chiamare to have to phone the teléfono de la esperanza para l'Associazione dei Samaritani e Samaritans and ask them. pedirles ayuda. Aunque, con chiedere a loro. Anche se, con la Although, knowing my luck, mi suerte, lo mismo me decían mia fortuna, mi sarebbe toccato they'd only give me lip or be cualquier grosería o me ascoltare il racconto della vita di completely distracted. ignoraban completamente. un Samaritano depresso.

Melanie is confused about her feelings for her boyfriend Alex and for her new friend

Angus. She tries to talk about it with her new work colleague, Jenny, but she is very upset because she had an argument with her boyfriend at the weekend and not being very helpful with Melanie. Melanie thinks that she would have to end up calling the Samaritans and asking for their help.

The name of this British association is being used to show the extent of Melanie’s feelings, she is so confused that considers she could enlist the help of the

Samaritans. This association has three tokens in the corpus, always used in a similar way, to illustrate, albeit in a somewhat exaggerated way, Melanie’s confusion and loneliness.

In the Spanish TT the Samaritans have been substituted with El Teléfono de la

Esperanza. This is a Spanish association which offers counselling and advice through the phone, in a similar way that the Samaritans do. The new name brings the novel closer to the Spanish culture, and it also evokes connotations of desperation and loneliness.

309

The Italian version of the novel literally translates the name of the association,

Samaritans – Samaritani, but adding some extra information, l’Associazione, to explain that the name refers to an association. As this charity does not operate in

Italy, there is a possibility that the Italian readers might not be familiar with the term and for some of them the connotations might be lost.

Example 13 British music band

(a) (b) SL Spanish Italian Amanda’s Procedure Procedure La boda de Non sposate Wedding Amanda quella donna! Runrig * Famoso grupo escocés de música Ø Runrig Gloss Omission folk-rock. (N de los T)

Concordance Line 13

English Spanish Italian Oye, Angus, ¿sabías que 'Hey, Angus, did you hear Runrig* va a hacer una versión Runrig are doing a Stones de un tema de los Rolling tribute cover?' Stones? Ø

The name of this band is part of a joke constructed using puns and homophones.

Some of the Scottish friends of Melanie are playing a joke on Mookie, one of

Amanda’s posh friends. She asks where they work, and they try to confuse her interchanging the oil rigs with Runrig, a famous Scottish rock band, known mainly in their country. The joke then carries on with a wordplay with the title of a song by the Rolling Stones and one by the Scottish band (example 19 below).

310

The TTs again show different approaches in Spanish and Italian. The first one uses a footnote to explain that Runrig is the name of a Scottish band. As they had previously translated the rigs as plataformas petrolíferas, the relationship between the band’s name and the job are missing and part of the joke is therefore lost. The

Italian translator, on the other hand, omits all reference to the Scottish band as well as all the jokes around the song by the Rolling Stones. The entire paragraph disappears from the Italian TT with no attempt to compensate for the loss in any way. This can be considered as an extreme case of cultural references treated as disposable elements. It is striking the extent to which the Italian translator prefers to omit an entire reference (and with it a key joke in the linguistic context of the novel) rather than introducing foreign elements into the TL culture or rather than attempting to explicate the references. These paragraph does not add any relevant material to the plot and therefore no major information which could hinder the understanding of the novel is missing from the Italian TTs. However, this joke recreates a sense of camaraderie between the characters, a joyful and relaxed atmosphere in the evening and these are lost for the Italian readers.

Example 14 UK Stereotype

(a) (b) SL Spanish Italian My Life on a Procedure Procedure Mi vida casi La mia vita su Plate perfecta un piatto Sloane Sloane Los pijos Generalization Ranger Gloss Ranger

311

Concordance Line 14

English Spanish Italian Sapete quella cosa per cui gli You know how the Sloane ¿Sabéis que, en teoría, los Sloane Ranger dovrebbero Ranger is supposed to be pijos han pasado de moda? essere morti e sepolti? Be', non è defunct? Well, it isn't. Bien, pues no. vera.

This British term refers to upper class women, and to some extent now also men, who share distinctive lifestyle traits and values, such as confidence in themselves and a belief that they are better. Clara uses this term to describe her rich step-father and his family.

The Spanish translator uses a generalization to describe the group simply as pijos

[posh]. The connotations evoked by this noun are very different to those evoked by

Sloane Ranger, and the specific referent to this stereotype in the British culture is lost. The Italian translator uses the glossary included at the end of the novel to add a definition of the term.

9.4.5 Pragnonyms

As seen in section 9.3 pragnonyms mainly refer to time period or festivities or events

The majority of these have an established translation in the target culture and therefore fixed translation stands out as the most commonly used procedure with this type of name. There are only 9 primarily connotative pragnonyms. The majority of them are irrealis names in the form of made up illness to stress some odd behaviour; these tend to be literally translated. Literal translation was the second also appeared as prominently used in this facet.

312

Example 15

(a) (b) SL Spanish Italian Amanda’s Procedure Procedure La boda de Non sposate Wedding Amanda quella donna! Unpopular at el síndrome la sindrome Literal Parties' de los sosos Literal translation da 'Sfigati Translation syndrome de la fiesta delle Feste (a) (b) SL Spanish Italian Bridget Jones’s Procedure Procedure El Diario de Il Diario di Diary Bridget Jones Bridget Jones Competitive Síndrome del Sindrome da Literal Literal Tiredness Cansancio Stanchezza Translation translation Syndrome Competitivo Competitiva Síndrome de sindrome del Having It All' Literal Literal «Tenerlo "Tutto e syndrome Translation Translation Todo» Subito"

There are also some realis, external names, with a referent outside the novel:

Example 16

(a) (b) SL Spanish Italian Amanda’s Procedure Procedure La boda de Non sposate Wedding Amanda quella donna! el mal de Colditz* * El Castillo medieval de Colditz, cerca de attacchi Colditz fever Leipzig, Alemania, fue utilizado por Gloss d'asma Generalization

los nazis en la II Guerra Mundial como prisión de máxima seguridad. (N. de los T)

313

Concordance Line 15

English Spanish Italian Slowly creeping my way off Me arrastré lentamente hasta el Strisciai verso l'estremità del the end of the futon, and extremo del futón y, sintiéndome futon, sentendomi sempre feeling worse and worse, I cada vez peor, fui a gatas hasta la peggio, e mi trascinai in cucina in crawled into the kitchen in cocina en busca de una aspirina y cerca di aspirina e coca dietetica. search of aspirin and Diet una Coca-Cola light. Fran, por Fran, naturalmente, era li che Coke. Fran, of course, was supuesto, aguardaba al acecho. No aspettava. Lei non viveva con lying in wait. She didn't live vivía conmigo, pero se sentía más a me, ma sembrava molto più a here, but she made herself gusto que yo en mi propia casa. suo agio di me in casa mia. La more at home than I did. Her Vivía en un estudio de un metro sola vista del suo monolocale di own place was a three foot cuadrado que te producía tre metri quadri provocava square studio which induced inmediatamente el mal de Colditz*, attacchi d'asma, perciò mi ero immediate Colditz fever, so de modo que me habia abituata a vederla entrare e I'd got used to her wandering acostumbrado a sus constantes uscire. in and out. entradas y salidas.

In this example there is more variation in the procedures used, with the Spanish translator introducing a footnote to explain the origin of the term and the Italian translator deleting the original name and all reference to it and introducing a more general term to describe an illness similar to those evoked by original.

Example 17

(a) (b) SL Spanish Italian The Bridget Procedure Procedure El Diario de Il Diario de Jones’s Diary Bridget Jones Bridget Jones Grasshopper la crisis de la cicala che ha Who Sang All Cigarra Que Fixed Fixed translation cantalo per Summer Cantó Todo el translation tutta l'estate crisis Verano

314

Concordance Line 16

English Spanish Italian Ore 14. Uffa. Proprio quello che 2 p.m. Humph. Just what I 2 p.m. Hum. Justo lo que mi ci voleva. Mi piombata in needed. My mother burst into necesitaba. Mi madre casa la mamma. La crisi della my flat, last week's irrumpió en mi piso, la crisis settimana scorsa, quella della Grasshopper Who Sang All de la Cigarra Que Cantó Todo cicala che ha cantato per tutta Summer crisis miraculously el Verano milagrosamente l'estate, l'aveva giá forgotten. olvidada. miracolosamente dimenticata.

Bridget uses this reference to the popular children tale of the grasshopper and the ant to describe her mother’s behaviour. Both translations use the versions of this tale in their cultures as the referent of their translations in an attempt to evoke the same connotations in the target readerships.

9.4.7 Ergonyms

Instances of man-made products used for their connotations are not uncommon and they encompass a variety of procedures for their translations; section 9.3 observed a tendency towards generalisation or literal translation with this type of name, the examples below show some of the uses of these procedures with connotative ergonyms.

Example 18

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! scena della (una escena Godfather (- testa di extraída de) El Fixed translation Explicitation type situation) cavallo del Padrino Padrino

315

Concordance Line 17

English Spanish Italian The chunk of flesh was El trozo de carne estaba conectado con Il pezzo era attaccato a connected to lots of other otro montón de trozos de carne, todos en svariati altri pezzi, tutti chunks, all in the right order, el orden correcto, pero no me percate de nell'ordine giusto, ma non but I didn't notice this until esto hasta que me in-corporé de un Salto ci feci caso finché non after I'd sat bolt upright in y me quedé tiesa, aterrorizada ante la balzai a sedere, col terrore terror at a potential posibilidad de encontrarme en una di trovarmi nella scena Godfather-type situation in escena extraída de El Padrino justo en mi della testa di cavallo del my bed. propia cama. Padrino.

This example takes us back to the beginning of Amanda’s Wedding, when Melanie wakes up in the morning and finds an unexpected man, Nicholas, in her bed. Before she remembers the events of the night before, she imagines in her head a

‘Godfather-type situation’. This refers to the popular film, and it implies associations with mafia and crime and has specific connotations related to certain scenes in the first film in the series.

The Spanish translator uses the established name of the film in Spain, el Padrino. As this is a very well-known film, the connotations associated with it can generally be expected to be understood internationally. The Italian translator also respects the established name for the film in the country, El Padrino. However, she also explicitates the meaning of the original, setting the scene clearly in relation to the scene in The Godfather I where they put the head of the horse in the bed of one of the characters and she wakes up in a fright surrounded by a puddle of blood. This was implied in the original and is spelt out very clearly in the Italian TT.

316

Example 19

SL Spanish Italian Amanda’s La boda de Procedure Non sposate Procedure Wedding Amanda quella donna! Oye, McLeod, bájate de mi oveja* ** La frase de Nash, Hey, McLeod, get off of my ewe, Hey, McLeod, hace alusión al Ø get off of my Gloss Omission título de una ewe canción de los Rolling Stones, Get off of my Cloud («Bájate de mi nube»). (N de los T)

Concordance Line 18

English Spanish Italian -Sí, ¡se va a llamar Oye, `Yes, it's called Hey, McLeod, McLeod, bájate de mi get off of my ewe!' oveja**! Ø

This is the second part of the joke introduced in example 13. Following with the homophones rig and Runrig, the Scottish friend casually comments that the Runrig are planning a cover version of the Stones called ‘Hey McLeod, get off of my ewe’.

This is a play on words trying to turn Scottish the real title of the song by the Rolling

Stones Get off of my Cloud. In the novel, only the Scottish characters understand the joke while Melanie, Mookie and the other people present do not.

317

The Spanish translators again introduce a footnote. The title of the Runrigs cover version of the song is translated in the body of the text and the footnote explains that it refers to the Stones song Get off of my cloud. To some extent this explains part of the joke, however, as the title of the cover version has been translated the phonetic symmetry of both titles is lost and the Scottish connotations of the pun are also missing.

As seen in example 13, the Italian translator omits the whole paragraph, and there is no reference at all in the Italian TT to this.

Example 20

SL (a) Procedure (b) Procedure My Life on a Spanish Italian Plate Mi vida casi La mia vita su perfecta un piatto Mills & Boon Al ratón y el Adaptation Mills & Boon Gloss gato

318

Concordance Line 19

English Spanish Italian

'I'd never hurt you.' 'Nor me E non ti farei mai del male. Né io you. Are we playing Mills & -Nunca te haría daño. -Ni yo a a te. Ma che, stiamo giocando a Boon?' I say, trying to lighten ti. ¿Estamos jugando al gato y Mills&Boons? dico io, cercando the oppressive atmosphere el ratón? -Pregunto para di risollevare l'opprimente that's suddenly gathered aliviar la atmósfera opresiva atmosfera che ci è calata around us. que de repente nos rodea. addosso all'improvviso.

Robert rimane ad aspettare, sulla Robert se para en la puerta porta, mentre accarezzo la Robert stands, framed in the mientras acaricio la cabecita testolina morbida di Jack, e door, while I stroke Jack's fluffy de Jack, y cuando levanto la quando alzo lo sguardo mi little head, and when I look up vista me sonríe con tal dulzura sorride con tanta dolcezza che, he smiles at me so sweetly que, si aún estuviéramos se stessimo ancora giocando a that, were we still playing Mills jugando al gato y el ratón, yo Mills&Boons, potrei dire & Boon, I might say something tal vez diría algo sobre qualcosa sul cuore e l'amore che about breaking and hearts. corazones partidos. va in frantumi.

This example is from the final part of My Life on a Plate. Clara and her husband have gone to Paris for the weekend and during dinner on the last day he announces that he is leaving her. They have a conversation about their relationship full of accusations and declarations of love. She feels as if they are ‘playing Mills & Boon’.

This is the name of a well-known publisher and alludes to romance novels with happy endings.

The Spanish translator adapts the name to the Spanish culture, where Mills & Boon novels are not published. There is no proper name in the Spanish TT, in its place the translator attempts to maintain the sense of the original, with a set Spanish phrase el gato y el ratón, which evokes a love-hate relationship with one always chasing the other. This fits well with the pace of the novel and what is happening in the context, but the connotations specifically associated with this very British romances are

319 absent in the translation. The Italian translator, as common for this language in this book, opts to borrow the term in the main body of the text while including a description of its referent in the glossary at the end.

9.4.8 Conclusions

These examples highlight a tendency to deal with connotations in a variety of ways, with different approaches from different translators or even from the same translators in the same book, depending on the name and to some extent also on the degree of familiarity with its referent and connotations in the TL cultures. In addition, a crucial point is also presented by the existence of a translation equivalent. If there is an established form for the term already in existence in the

TL, it can be presupposed that the TL readers could also be familiar with the connotations associated to the referent; changing the name or maintaining it in the

SL (as was the case with example 9 - Malory Towers) brings about a loss of intertextual cohesion in the novel.

Borrowing and fixed translation have transpired as by far the most used procedures for names with referents also known or famous in the target cultures, this is the case, for instance, of the Godfather in example 18 or Doris Day in example 4.

A degree of subjectivity is of course always present, as is always the case when dealing with connotations. This is true even for the SL audience; authors can never guarantee that their readership is going to perceive from a name exactly what they want them to perceive, up to some extent their ability to pick universal enough

320 referents, with very specific and recognisable characteristics associated to them, contributes to the success of the novels (as described in Chapter 5). The descriptive toponyms in Example 10 are a very good illustration of this – the subtle difference between Kensington and Kennington might be more obvious for a Londoner than for a person from any other part of the UK, whereas the traits associated with the

Cotswolds might reach a wider audience. This is also true for the TTs, and translators need to balance the smooth running of the text, or indeed a desire for domesticating naturalness, with a guess of what could be understood or accepted in their target cultures. Gutt’s (1998) application of relevance theory to translation was briefly introduced in Chapter 2 in relation to the extent to which the situation in which a proper name is used in a linguistic exchange determines the referent of that particular name and how this is crucial for the translation setting. The same applies here, as Gutt explains the context in the mind of the author might not always be the same to that of the audience even in the SL, and it is always a challenge for the translator to establish that context in the TL culture (Gutt, 1998). A very illustrative example of this is the treatment that the footballer David Beckham [Anthroponym/ famous figure/ primarily denotative/ realis/ external] receives in the Spanish version of My Life in a Plate. The scene is very simple, Clara’s youngest son is talking to his step-grandfather and telling him that his favourite sport person is David Beckham.

The Spanish translation, published in 2001, includes an extratextual gloss, a footnote, to explain that he is a famous footballer who plays for Manchester United.

The translator felt the need to guide the readers with this name. However, a few years later this would have been totally unnecessary, as David Beckham moved to

Real Madrid and everyone in Spain was by then familiar with him. This example also

321 very clearly illustrates the dangers of overexplicitation, as the footnote seems redundant as well as obsolete for today’s readers of the novel as Beckham has long been gone from Manchester.

On the opposite extreme lays example 13, Runrig, obscure even for SL readers, in fact not even the English characters in the novel understand the joke that the

Scottish blokes are laughing away at. In this case, the Italian translator took the decision to omit the complete paragraph rather than introducing unfamiliar, very foreign elements into the TL culture.

Notwithstanding all these, the tentative generalization that can be extracted from the study of all these connotative names in context, is that to some extent the procedures around the substitution extreme of the continuum seem to contribute to reflect or transmit some connotations, whereas the procedures around the conservation extreme introduce foreign elements into the TL culture, and leave entirely to the reader the interpretation of the message portrayed by the referent and its connotations.

This qualitative analysis has also been directly linked to the patterns explored in section 9.3. First of all, these examples of names in context highlight the wide variety of procedures and processes used with connotative names. As seen in section 9.3, the semantic facet and more specifically the primarily connotative names registered the highest volume of variation between the patterns of usage recorded at procedure level and those recorded linking this particular type to the procedures

322 used to translate this type of name. Possible reasons for this have been explored above. With regards to the other categories, the names analysed in context illustrate the trends observed for each category where this differed from the norm found at procedure level. Overall, this section explored more in detail the possible reasons behind the choice of procedure by the translator.

9.5 Conclusions

This chapter collects and interprets all the figures from the study of the classifications of the proper names in BRIDGET and the procedures employed for their translations into the TLs. It starts with a global overview of the data across all

TTs. An initial pattern was visible even from this broad approach to the figures: borrowing immediately stands out as the most frequently used procedure overall. A more detailed look at the figures followed, with a close study of the procedures employed in each book. Broadly speaking, this points towards a wider use of procedures around the conservation extreme of the continuum in the Spanish TTs as opposed to a wider use of the procedures around the substitution extreme in the

Italian TTs. Thirdly, these figures on the frequency of use of each procedure were combined with the data extracted from the classification of proper names according to the taxonomy devised for this study (Chapter 7). From this exercise a relationship ultimately emerged between procedures usage and name category, with a clear tendency for more substitution happening around chrematonyms, ergonyms, and primarily connotative names. This led to a more qualitative exploration of the names under these categories, specifically around the primarily connotative names, where

323 the most variation was detected. The sample of examples hand-chosen from all the connotative names in BRIDGET showed variety in the way that translators deal with connotations, depending upon both the context and the text as well as upon the referent and what it evokes but ultimately mainly upon the TL culture and their acceptability of the new name.

The bullet points below concretely set out all the findings:

 Borrowing, fixed translation and literal translation consistently have the

highest relative frequency across all TTs covering a total of 80% of all the

translation instances

 Higher relative frequency of the procedures around the substitution

extreme of the continuum in the Italian TTs than in the Spanish TTs

 The general pattern for the translation of proper names in Chick Lit

seems to show a mixture between conserving the text with foreign elements

and substituting some others

 An exploration of translation procedures and taxonomy for proper

names show that chrematonyms, ergonyms and primarily connotative names

are the categories that tend more to be substituted over any other category

 Ergonyms tend towards generalization

 Pragnonyms and toponyms tend towards fixed translation

 Chrematonyms show a wide range of procedures used for their

translation

324

 Names classified as anthroponym/nickname/irrealis/internal tend to be

adapted as TL nicknames

 Primarily connotative names tend to be substituted more

 Anthroponyms are far more likely to be used for their connotations

than any other category

 Translators deal with connotations in a variety of ways – this links

directly with the general pattern showing a mixture between conserving and

substituting

 Borrowing and fixed translation tend to be used with names where the

referent (and therefore to some extent also its connotations) are also known

in the TL cultures

 Procedures around the conservation extreme leave it up to the reader

to understand the referents and connotations

 Procedures around the substitution extreme attempt to decipher the

connotations for the TL readership

Some of these findings may just state the obvious, for instance it can be expected that borrowing ranked as the most prominent procedure especially taking into account the general belief in TS of proper names never being translated (Chapter 3).

However, crucially, this study offers empirical evidence to actually show this point.

In contrast, another interesting general pattern highlighted by this analysis is the differences about the translation of core versus peripheral terms, in other words, internationally well-known references tend to be transferred across in their established forms in the TLs and in this way their associated connotations travel with

325 them. On the other hand, lesser known references tend to be subjected to more varied treatments in the translations. Some of the connotations or allusions are modified to the perceived requirements of the TT audience. The concepts of translation variation and translators choices are paramount here. These highlight the differences that can be appreciated in the figures between the uses of fixed translations and possible variations (for instance variations with a default option or variation without a default option). As shown by the figures, fixed translation tends to be used for proper names that have a form already established in the target language culture. As a general rule translators tend to respect this form. This is the case of names of countries, cities or towns: London/Londres/Londra. There are other cases, though, where the division is not so clear and then translation variations occur. An example of a variation with a default option, for instance, is the case of the translation of the title of the TV programme Blind Date. This programme features regularly as one of the preferred options for the characters to watch on

Saturday night. Different translations of this ergonym occur in the books. The default option would be to find the equivalent on the target languages, using the procedure adaptation (Blind Date/El flechazo/Appuntamento al buio); however this option is nor always used neither obligatory, and sometimes a borrowing is deemed by the translator more suitable to transfer the name across (for instance, with the aim to avoiding moving the setting of the novel to the TL culture). On the other hand, there are also cases where translation variation can be appreciated where there is not a default option. Honorific titles are an example of this kind of variation, especially to transfer titles that do not have an equivalent in the target cultures. The most illustrative example of this particular kind of variation is the different options

326 taken to translate the Scottish title Laird (that Fraser inherited from his grandfather).

In one instance, the Spanish translator uses a gloss to explain that this is a Scottish title similar to a Lord in England; in other cases the translation is Lord. Similarly, when this title is applied to Amanda (Lairdess Amanda) the Italian translator borrows the term, but the Spanish one changes it to simply Señora, omitting the connotations of the original. These examples of translation variations with or without a default option contribute to some of the fluctuations that can be appreciated in the figures. All options are correct and explored by the translators, who always have to take choices.

These variations explain why the concrete figures set clearly and comprehensively a trend in the translation of proper names that, to some extent differs from that rule of proper names not changing in translation. Despite the fact that borrowing constantly ranks as the most commonly used procedure, variation, changes and shifts are also very visible and real in all the TTs, above all in the connotative names identified in BRIDGET. From that standpoint, in the discussion that follows (Chapter

10), all these findings are put in relation to the general theories of TS (introduced in

Chapter 2) as well as more concretely with previous studies on the translation of proper names (Chapters 3 and 4).

327

Chapter 10. Discussion of the Results

“Call him Voldemort, Harry. Always use the proper name for things. Fear of a name increases fear of the thing itself”. Rowling, J.K.: Harry Potter and the Philosopher Stone (p. 216)

10.1 Introduction

In Chapter 9 I presented detailed results from the analysis of the proper names extracted from BRIDGET and their translations into Spanish and Italian and concluded with findings which can be inferred from the interpretation of these results. In the present chapter, I will place these findings, both concrete and general, in the context of the theories reviewed throughout this study. I will discuss the results in light of Toury’s (1995) laws and norms, as well as in relation to the concept of translation universal (Baker, 1993, Chesterman, 2004) and of foreignization and domestication (Venuti, 1995/2008). The theories of meaning and the general theories of proper names (Mill 1843, Frege 1892, Russell 1905, Searle

1958, Kripke, 1970/1980) together with the theories of the translation of proper names (Nord, 2003, Moya 2000, Cuéllar Lázaro 2000, Franco Aixelá, 2000,

Leppihalme, 1997), will also play a role in this discussion.

10.2 Exploring the relationship of the findings to TS

At the beginning of this study [Introduction and Chapter 1] I set out to identify the general pattern for the translation of proper names in Chick Lit novels using Toury’s

328

(1995) methodology for DTS. From the standpoint of this methodology, comparing

SL with TL adjacent pairs, it should be possible to identify patterns of translation behaviour and thus to express generalizations about what happens during the translation process, that is, from ST to TT. Toury’s (ibid.) concept of norms allows for these generalizations to be expressed and categorised. The relationship between

ST and TT segments was explored in terms of the translation procedures used with each segment.

In Chapter 8 I placed translation procedures on a continuum portraying two leading tendencies, be it to conserve the foreign name and hence also its referent or to substitute it for an alternative. This directly relates to Toury’s initial norm (ibid.) setting the text either towards SL or TL orientation. Thus, the procedures around the conservation extreme of the continuum show a tendency towards adequacy (SL interference – ST oriented) and the procedures on the opposite extreme of the continuum, the substitution extreme, show a tendency towards acceptability

(following TL rules – TT oriented).

329

Figure 10.22: Continuum set against Toury’s initial norm

The results from the analysis of the procedures used with the names in BRIDGET seen in Chapter 9 clearly demonstrated how the Italian texts consistently showed higher frequencies for the procedures around the substitution extreme than the

Spanish texts, which had consistently lower frequencies in these procedures. In

Toury’s terms, these findings seem to show a tendency towards adequacy in Spanish texts in contrast with a more marked tendency towards acceptability in Italian texts.

The exploration of proper names in context presented in Section 9.4 clearly evidences how the procedures around the conservation extreme leave it to the reader to understand the referents whereas the procedures around the substitution extreme help guide the reader in deciphering the referents and their connotations. A similar point was also shown by Leppihalme (1997) (section 4.3.4) through the interpretation of reader response to translated proper names. With those procedures

330 which involved more extreme changes to the referents, the readers reported fewer problems in interpreting the meaning. This fact has also been highlighted in this study and it can be very clearly visualised with some examples from the corpus, for instance, contrasting Example 9 with Example 10 of the names analysed in section

9.4. In example 9 (Malory towers) both translators, Spanish and Italian, use a procedure from the substitution extreme of the continuum modifying to an extent the referent and its connotations for their respective audiences. In contrast, the names in example 10 (Kennington and Kensington and Cotswolds and Crouch End) do not vary and the reader has to be familiar with the regions being referred to in order to fully understand the authors’ meaning centred on the differences between the places. So from these findings, it can be inferred that Italian translators more consciously try to bring the text closer to their TL readership than the Spanish translators, who show a clearer tendency to adhere to the SL more closely. It has been common to take this approach to mean that the texts are either ‘domesticated’ or ‘foreignized’, using Venuti’s (1995/2008) terms. Davies (2003) is one author who argues against this belief. In Section 8.3 we addressed Davies’ objections to Franco

Aixelá’s gradation between the procedures. In her article, Davies also claims that there is not ‘a predictable correlation between the degree of manipulation of the ST and the extent to which the TT is domesticated’ (ibid:97); in her opinion some changes brought to the TT may still contribute to making the text SL oriented, for instance with the replacement of certain products for other products still foreign in the TL culture. Below, I would argue with Davies’ ideas that the use of one procedure or another does not contribute towards making the text more clear for the

TL readers, but I would agree with her objection to using the concepts of

331 domestication or foreignization in the same fashion. Although I use Toury’s concepts of adequacy and acceptability (Toury, 1995) to pinpoint the general orientation of the translation, this does not necessarily imply domestication or foreignization, as I understand Venuti’s concepts. Similarly, Davies argues that manipulation of ST does not necessarily equal domestication (2003.:79-82); for example, she believes that the use of omission can sometimes ‘make the text more accessible’ (ibid:97); and indeed this might be the case. To use one example from BRIDGET, in the novel

Amanda’s Wedding both the Spanish translator and the Italian translator choose to omit the name of Asprey’s bar where the characters meet one night; in the TTs they simply meet at a bar. This avoids the introduction of a referent which might be unfamiliar for their readerships, and can be taken as a move towards making the text clearer for them; indeed, this procedure has been placed in the far right corner of the substitution extreme of our continuum. However, that is not always the case:

Marco (2004), for example, leaves omission out of his continuum as he finds it difficult to identify the right place for it (section 4.3.5).

Still, controversy of where to place omission aside, I would argue with Davies that a change in referent, or the consistent preference towards the procedures on the substitution extreme of the continuum, has to be motivated by an intention to make the text clearer, less obscure for the TL readers. It can, nonetheless, of course be the case that despite many changes being applied to the text it stills contains its foreignness, the key lays, no doubt, on the nature of the procedures used. In other words, the use of cultural substitution to replace a foreign name with another foreign name (for example, replacing Bagpuss with Gatto Silvestro) contributes to

332 maintaining the foreignness of the text more than the use of cultural replacement where the name is replaced by a name from the TL culture (e.g. ER / I casi della vita), but still the intention of both is to bring the text closer to the TL readers, as

Marco’s (ibid.) continuum suggests.

Thus, a move towards acceptability in this sense is not necessarily equivalent to a move to domesticating the text. Venuti (1995/2008) conceives the concepts of domestication and foreignization in the context of translation from minority cultures to dominating cultures, where the minority culture is to an extent suppressed in translation. Furthermore, he later (Venuti, 2008) sees the concepts more as translation methods adopted by the translator when faced with the task of introducing a new text into the TL culture. This does not imply, however, that the individual procedures have a radical domesticating or a radical foreignizing effect.

The fact that the Italian translations in BRIDGET tend towards acceptability does not mean that these texts are deprived of all elements of the SL. In fact, in many examples the procedure used simply replaces the foreign referent with another foreign one, albeit one presumably better known in the Italian culture (for instance, replacing Miriam Margoyles with Nonna Pappera (Donald Duck’s grandmother) or

Anthea Turner with Kate Moss). By the same token, the fact that the Spanish texts show a more marked interference from the SL does not mean that these texts are completely foreign for the Spanish reader unfamiliar with the British culture. For this reason, I see a clear correlation between Toury’s concepts and our continuum, and take adequacy and acceptability as descriptive categories to identify translation patterns, but do not necessarily see the same correlation with Venuti’s concepts. In

333 this sense, these results set Italian translations towards acceptability and Spanish translations towards adequacy but not necessarily making the texts more foreignized or domesticated.

The second most prominent finding in Chapter 9 is without a question the high level of usage of repetition, followed by fixed translation and literal translation. This matches findings identified in other studies on proper names: Moya (2000), for instance, documents how the tendency in Spanish newspapers at the end of the twentieth century was to maintain the foreign names. This again supports a preference towards adequacy in Spanish translations.

Toury himself stresses (1978/2004) the fact that phenomena that occur with high frequency may be taken as evidence of ‘universals of translation behaviour’ and as such, perhaps it could be generalized that these procedures represent the general pattern for the translation of proper names. Indeed, this feature of proper names can be linked to that ‘rule deeply rooted in many people’s minds’ (Nord, 2003:182) that proper names never change in translation. As seen in Chapter 3, many theorists of translation argue in favour of the impossibility of the translation of proper names

(notably, Zabeeh (1968) and Manczak, (1991)). Taking Universals to ‘represent general tendencies and [be] observed irrespective of the translator, language, genre or period’ (Kenny, 2001:52) these figures, which occur across all TTs in both languages of this study, could be taken as a further contribution to support this idea with a relevant amount of new data as well as empirical evidence and statistics. All this could potentially be converted into the universal that proper names are not

334 modified in translation. However, the term universal seems to have been taken loosely in this matter; despite the fact that the first thing that comes to many people’s minds when mentioning the translation of proper names is that they never change in translation, the reality that has been shown with the examples in contexts used throughout this study is that proper names play indeed a very important role in translation and, as Hermans points out they ‘constitute a privilege object of enquiry’

(Hermans, 1988:14). Hence more than a universal of translation this seems to validate a generally-held belief about the translation of proper names. Furthermore, in contrast with the idea of proper names not changing, the data from this study clearly stresses how although the vast majority of the evidence points towards conserving the foreign names, there can also be found a non-negligible amount of usage of alternative procedures, all along the continuum, a fact that contributes to further reject the idea of proper names not changing in translation as a universal. I agree with Chesterman (2010) and Hermans (1999) that the so-called universals of translation can never be really proven to be completely universal, as it is not feasible to study all translations available (Chesterman, 2010:44) or to take into account all possible variables (Hermans, 1999:92); moreover, the data from BRIDGET shows that the remaining 20% of names do not agree with this belief as they behave differently and indeed change in translation. Therefore I will be staying clear of the quest for translation universals on proper names. Nevertheless, this does not devalue the fact that up to 80% of all the translation instances were transferred into

Spanish or Italian through the use of repetition, fixed translation or literal translation. This forms a very useful and powerful finding for scholars working in machine translation. This data validates the idea that perhaps the translation of

335 proper names can be semi-automated with these three procedures, ensuring a turnover of 80% accuracy in the results obtained.

In terms of Toury’s laws (Toury, 1995), the opening stages of this study

(Introduction) saw the formulation of the hypothesis that the law of growing standardization would prevail, as translators attempt to bring the connotations and foreign elements of the SL text nearer to the TL readership. The high level of use of procedures on the conservation extreme of the continuum indicates a rejection of this hypothesis. However, the data from this study strongly points towards the use with certain names of alternative procedures, those on the opposite extreme of the continuum, which favour structures more common in the TL. Thus, in reality, the pattern that this data seems to reveal for the translation of proper names in Chick Lit is a tendency towards a mixture between the conservation of some of the referents, manifesting interference from SL, and the substitution of some others, making the text more standard and typical with the structures of the TL. Therefore, in summary, it cannot be affirmed that the pattern for the translation of proper names follows solely the law of growing standardization, nor can it be stated that it strongly adheres to the law of interference. Munday (2012:179), talks about the limitations of

Toury’s laws, contemplating the possibility of adding a ‘third law of reduced controlled over linguistic realization in translation’ (ibid.) and perhaps the problems with anchoring the results from this study to either one of the laws proposed by

Toury can be used to further strengthen this point. Munday explains that translation is in reality ruled by ‘constraining factors’ (ibid.) which include for example ‘the preference for clarity and avoidance of ambiguity in TTs and real-life considerations

336 for the translator’ such as ‘efficiency of thought-processes’ and ‘time pressures’

(ibid.). Results from other studies described throughout this project seem to support this idea. For example, Franco Aixelá (Franco Aixelá, 2000) concludes that the TT had to be adapted to the TL culture in order to make it viable from a commercial point of view. Leppihalme’s (1997) empirical work with readers of the translated texts again stresses how sometimes information that is left unchanged or unexplained is lost for the target audiences. It all seems to point, therefore, more to a need of translators to balance different social and linguistic factors. Pym (2008) maintains that both of Toury’s laws could be explained by translators’ desire to avoid risk, Chesterman (2010) succinctly summarizes Pym’s ideas as follows:

Translating literally (which might involve interference) is a way of playing safe if you are not sure of the exact meaning of the source; and using high- frequency forms is also a way of playing safe not only from the point of view of getting the message across to a wide readership, but also in regard to minimizing the risk of unnatural language. (Ibid.:43).

These factors can all be strongly applied to Chick Lit. As seen in the description of this genre included in Chapter 5, these novels are taken as light reading and readers are not expecting to be challenged by the text. On the contrary, they are mainly searching for a good, escapist time, but equally, the very British character of the books cannot be overlooked. This puts the translator in the difficult position of having to balance an easy read with the necessity of making the text far reaching, ensuring it is understood by as wide an audience as possible in the target culture. A

Chick Lit text full of obscure references, inaccessible for the average reader, will not

337 work in the commercial market where these novels operate. By the same token, a

Chick Lit text robbed of all foreign referents would stand out from what is expected from the genre. Mayoral (1994) joins voices with Leppihalme (1997) in highlighting the important role that the target audience plays in the choices that a translator makes:

las razones para mantener un máximo de color local y temporal, un máximo de caracterización cultural son principalmente razones de estilo. Las razones para reducir su proporción son razones de compresión por parte del lector. El punto de equilibrio lo establece el lector al que nos dirigimos. (Mayoral, 1994:88).

The findings from this study seem to represent very clearly all these ‘constraining factors’ (Munday, 2012:179) and the search for ‘equilibrium’ (Mayoral, 1994:88) that the translator undertakes. The majority of the names are in fact kept unchanged, or adapted to the standards of the TL and its culture with a fixed or literal translation; but, on balance, a great number of names are also substituted and adapted. This is done in order to make the text viable, in an attempt to transfer to the target culture and to the TL readers the cultural background and sense of place evoked by proper names for SL readers. Of course, in the search for this balance, the translator has to take choices; s/he has to decide what to change and what to leave intact.

The exploration of translation procedures and taxonomy for proper names undertaken in 9.2 of this study highlights the type of names that tend to be more substituted in these translations. Primarily, within the ontological facet of the taxonomy, the results show that chrematonyms (names of organizations) and

338 ergonyms (product names) show a more marked variation; with anthroponyms

(names of individuals) showing the most variation of all types. A closer analysis shows that ergonyms tend towards generalization more than any other procedure.

This means that product placement and localization play an important role in these novels with regards to the names of the brands and things that the characters consume. Likewise, chrematonyms stand out for being the type of name showing the most widespread use of procedures across the continuum. This can be explained by the nature of the category, which incorporates a wide variety of entities (the chrematonym hypertype contains a total of 10 distinct types, a fact which makes it the second most varied hypertype in the taxonomy behind only toponyms) which are, at the same time, dealt with by the use of a wide variety of procedures.

On the other hand, and most importantly, the combination of taxonomy and procedure usage more markedly highlights two main points: firstly, that connotative names tend to be substituted more than any other category, and secondly that anthroponyms are the kind of name most likely to be used for their connotations.

The Chick Lit genre uses real names to evoke and recreate the world where both character and readers live (see Chapter 5); this study shows that anthroponyms are the preferred means to recreate this familiarity for the audience. This was seen in the examples used throughout this study and more specifically in 9.4 with the analysis of examples in context, where the majority of the names selected analysed where indeed anthroponyms (eight examples out of 20 were from this category, with the other 12 examples being shared by the remaining four hypertypes in the taxonomy). We see how Alice Cooper is chosen to describe the character’s eyes, or

339

Paddington Bear to recreate a mean stare. James Bond is used to illustrate a type of trousers or Doris Day a way of being. This, again, correlates with the more general findings of this study, the fact that translators have to balance the foreignness of the novel with making it understandable for their target readers and the results show how the Italian translators tend more towards substituting the names than the

Spanish translators.

This once again highlights the relationship between the theories of meaning and the translation of proper names. As seen in Chapter 3, until recently proper names were taken as individual objects, belonging mainly to the ‘encyclopaedia and not to the dictionary’ (Newmark, 1981:70), and they were not considered for translation. But if we take into account the theories of meaning discussed in Chapter 2 (Frege (1892),

Russell (1905 and 1912), Kripke (1970/1980), Searle (1967)), and see proper names as connotative objects in their own right and context, then they become one of the more specific problems of translation. This study shows proper names as connotative objects, used mainly for their power to evoke certain ideas in the readers. The analysis of the procedures used for their translation highlights that translators deal with connotations in a variety of ways, maybe not always consistently, as they have to balance what they consider to be important for the plot and what they consider may need an explanation for their audiences. What most prominently has been highlighted by this study is the different factors at play when it comes to transmitting the contents of a proper name to a new target audience, and the necessity to take each individual name in its context as well as all the referents as a whole in order to make sure that the equilibrium is maintained between making the

340 text understandable and ensuring it remains full of the SL culture. This can be linked with more ground-breaking approaches being nowadays proposed in CBTS as the way forward (1.4.1). Scholars such as Laviosa (2011), Munday (2001), House (2011) and Saldanha (2011) are now calling towards a use of CBTS as a means to study extra-textual aspects of translations. This study sheds new evidence towards this necessary move in the discipline, as the way to exploring both the real potential of a corpus approach as well as the means to fully explaining what really takes place between ST and TT.

10.3 Conclusions

In this chapter I have discussed the relationship of the findings from this project to the theories reviewed throughout the present study. Overall, it can be affirmed that the general pattern for the translation of proper names in Chick Lit follows a mixture for conserving certain elements together with substituting some others. It has been shown how the target audience, as well as other ‘constraining factors’ (Munday,

2012:179), plays an important role in managing the balancing act that translators perform, between making the message clear and maintaining the foreignness of the novel. In the search for this balance, the results from this study also evidence that the Italian translators tend towards ‘acceptability’, substituting more elements, than the Spanish translators who more markedly tend towards ‘adequacy’, remaining closer to the SL (using Toury’s initial norm concepts (1995) for describing translators behaviour). Ultimately, the need for new approaches into CBTS research

341 incorporating (House, 2011, Saldanha, 2011) extra-textual factors to their analysis has been prominently highlighted as the way forward.

In the next chapter I will return to address the aims and objectives of this study describing concretely where they have been met. I will also consider the limitations of the project within its scope and point to future areas for research.

342

Chapter 11. Conclusions

He had stolen the name from a history book in his old wold. A famous knight once bore it, a king’s son who was a great robber too. Would he have been pleased to think that his name had been given to a knife thrower, king of the strolling players? Funke, C.: Inkspell (p.153)

11.1 Introduction

The present project is distinctive in systematically studying proper names in the translation of Chick Lit from a cultural and functional point of view. Traditionally, proper names have been considered independently from the rest of the text

(referring back to the theories of the impossibility of the translation of proper names explored in Section 3.2, Zabeeh 1968, for instance). However, this study demonstrates that proper names convey different levels of information and play specific roles depending on the context and the function sought by the author, and come to underpin the importance of taking into account these as well as other external elements when transmitting those connotations to the target readers.

Critically, this point links the findings from this study to the latest trends in CBTS, where similar calls for inclusion of external elements into the corpus-based approach have been worded by House (2011), Munday (2011) and Saldanha (2011).

In addition, the methods of analysis devised and successfully applied in this research represent the lasting contribution of this study. The taxonomy for proper names and the map of translation procedures can now be used in other investigations and the

343 methods employed to quantify the uses of the different categories in them represent a theoretical advance that can also be replicated in other studies.

I will start these concluding remarks with a brief summary of the work undertaken and a short analysis of the positive and negative aspects faced along the way. I will then address the aims and objectives, pointing out where specifically they have been met. I will also assess the extent to which the research questions have been answered. Lastly, I will talk about the limitations of this study and focus on aspects that could be considered in future studies.

11.2 Summary of the work undertaken

In this thesis I have investigated the treatment that proper names received in the translations into Spanish and Italian of three British Chick Lit novels in order to pinpoint the tendencies and patterns for the translation of proper names in modern literature.

I first highlighted the relationship between the theories of meaning and the theories of the translation of proper names and worded a definition of the term from these theories. The more practical work saw the creation of a parallel corpus of British

Chick Lit novels and their translations into Spanish and Italian. This corpus was used as the main source of the data. With the aid of corpus tools, such as named entity recognition programs and a multilingual concordancer, and computerized measurements such as mathematical functions in Microsoft Excel, I then undertook a

344 systematic analysis of all the proper names in the corpus. I devised taxonomy for proper names and a map of translation procedures and applied them to the list of names and translation instances extracted from the corpus.

This data acquisition and analysis process had positive and negative moments. The creation of the corpus was a slow, time-consuming task (in particular, the aligning of the paragraphs), but in contrast, once the corpus was put together and was ready to be interrogated it meant that a very useful resource had been created, for use in this project as well as in future ones. Similarly, classifying each proper name, both from

SL texts and TL texts and each translation instance was also a lengthy task that required great effort to ensure consistency. Once all the data was ready, the analysis and information that I was able to infer from it was again very rewarding. Using different metrics and formulas I devised a method for accounting for the use of the translation procedures. Generating data on the raw and relative frequencies of the use of each category within the maps and plotting these figures against a continuum

I was able to use Toury’s concepts of acceptability and adequacy to interpret the patterns shown by the data and successfully pinpoint the tendency behind the translation of proper names in Chick Lit, thus meeting the aims of this study.

11.3 Aims and objectives

The general aim of this research was to offer an overview of the role and treatment that proper names receive in Chick Lit and its translation and was successful at highlighting a tendency towards adequacy in Spanish translations and a tendency

345 towards acceptability in Italian translations. More importantly, this research stressed

(Chapters 9 and 10) the attempts of translators to identify an ‘equilibrium’ (Mayoral,

1999/2000) between conserving certain elements and explaining others. The search for this equilibrium involves different elements, such as the referent behind the name and the intention sought with it, the target readers and their (supposed) familiarity with the source culture, as well as other external elements such as the editors, time-pressures and similar constraints. These empirical pieces of evidence point towards the benefits that incorporating these external features into the translation research could bring to the discipline.

With regards to the concrete outputs set out in point 5.1 of the Introduction, I will comment individually on each one of them below:

1. To construct a theoretical framework and elaborate a definition

for the term ‘proper name’ which could link the most salient

theories of meaning to the theories of the translation of proper

names and could be appropriately used for the investigation of

the treatment of proper names in the corpus;

The following definition for proper name was given in Chapter 2:

‘Proper name: a word, or group of words, that identifies an individual entity

setting it apart from all the other entities in the same class and which, in

the context and situation in which it is being used, has a specific referent

346

and sense. Proper names both denote this referent and connote the

characteristics which form its sense’

This definition takes into account linguistic and, more importantly, semantic attributes of the term. The most salient feature highlighted in this definition is the importance of the referent behind each individual use of a proper name. This point was first introduced by Jespersen (1924/1964), but seems to have been neglected in the majority of the literature on the subject. Theories of meaning have been used by scholars interested in the translation of proper names (for instance, Allerton, 1987 or

Pendelbury, 1990) to argue the possibility or impossibility of the translation of proper names. Chapter 3 looked into the relationship between these two disciplines, and again implicitly highlighted the importance of the individual referent for the translator (for instance, the example used in Chapter 2 of a reference to the character Paddington Bear as opposed to a reference to the toddler’s teddy called

Paddington. The referent behind each occurrence of the name is different as are the connotations associated with each one of them, and the translator needs to be aware of this factor). Scholars such as Martinet (1982) or Cuéllar Lázaro (2000) stress that proper names with semantic meaning should be considered for translation and with this research I intended to go even further to show that in order to determine whether a name has semantic meaning the specific referent is paramount. Thus, the individual referent becomes a remarkably prescient concept and as such has an instrumental impact in the practical components of this research.

347

2. To create a parallel corpus of Chick Lit novels to be used as the

main source of the data for the investigation;

This objective was met as the first step in the more practical and applied parts of this research. The corpus, hosted by the University of Leeds, has been called

BRIDGET and it currently contains 652,926 words in total. It includes three original

British novels and their translations into Spanish and Italian. BRIDGET is aligned at the paragraph level and, although originally envisaged as raw text, it was subsequently tagged in order to generate word-frequency lists. Chapter 6 presents the specifics of the design and creation of BRIDGET.

3. To automate the manipulation of the corpus in order to

achieve maximum efficiency and accuracy through an automatic

extraction of the proper names in SL texts as well as their

translations in the TL texts;

Meeting this objective was more problematic than anticipated. I originally envisaged a straightforward use of an already existing named-entity recognition program

(GATE) to identify and extract all the names in the SL texts in BRIDGET. This, however, returned a lower than expected recall and alternative methods had to be considered. The reasons for this low recall have their roots in the high numbers of connotative names in BRIDGET. These names are used primarily for their attributes as opposed to as referential entities, and do not always conform to a set grammatical rules which can be automated by a computer. Ultimately, it was a

348 combination of several different approaches which contributed to the almost 99% recall finally achieved. Chapter 6 describes all the experiments which took place. The whole list of proper names, and their translations (successfully obtained through the use of the multilingual concordancer), manifest the concrete output of all this empirical work. This list is included in the attached CD in Appendix 2 for Amanda’s

Wedding, Appendix 3 for My Life on a Plate and Appendix 4 for The Diary of Bridget

Jones.

On the other hand, the inefficacy with the existing technology also required the creation of a program specifically designed to extract proper names from texts. This program, Access for BRIDGET, becomes, thus, an extra output achieved by this project. After initial teething problems the program was used successfully to achieve maximum recall in the results obtained. It will now be available to use with other texts and in other projects.

4. To design and practically apply a taxonomy for proper

names to classify and structure the data extracted from the

corpus;

This taxonomy was created taking into account latest developments in information science as well as work around the creation of an online dictionary of proper names which is being conducted by the PROLEX group (Grass, 2000). The faceted classification and the processes used for its application with BRIDGET are described in Chapter 7. The concrete output which sees the achievement of this objective is

349 again collected in Appendixes 2, 3 and 4, where a classification of all the names, both SL names and TL names, can be viewed. Pick lists and filters allow for the information to be viewed by facet or category, any combination is possible, and this is the main benefit of adopting a faceted approach to classifying the names.

This taxonomy and the processes devised to apply it to the names extracted from

BRIDGET represent one of the major research results of this study. This taxonomy is ready and can be applied to other large-scale investigations.

5. To design and apply a map of translation procedures for

proper names to pinpoint the techniques employed to transfer

proper names from SL texts to TL texts;

This map is presented in Chapter 8. It integrates different maps which have been created specifically to classify the procedures used with culture-specific elements.

These maps (Newmark, 1988, Hermans, 1988, Franco Aixelá, 2000, Leppihalme,

1997 and Marco, 2004) were first introduced in the theoretical parts of this thesis, on Chapter 4.

The model combines an important number of previous attempts at mapping procedures for culture-specific elements; it takes its inspiration from Marco’s model and uses a continuum to indicate the gradation between the procedures. In this project it was successfully used to interpret the results obtained (Chapter 9) and thus represents another important achievement. The key parameter introduced

350 again relates to the individual referent behind each name, as this is taken as the criterion for locating the procedures along the continuum. Each individual extreme of the continuum takes opposed tendencies. The gradation between the elements along the line is less marked but, with the change in referent as the main criterion, still palpable. For instance, a literal translation conserves the same referent whereas a cultural substitution changes it by a different one, consequently, literal translation takes a place nearer the conservation extreme of the continuum and cultural substitution one nearer the substitution one.

Chapter 8 also includes an account of the processes used to apply this map to the translation instances from BRIDGET. The full classification can be viewed in

Appendices 2, 3 and 4.

6. To generate and interpret data on the frequency of use

of each procedure alone and in combination with the categories

in the taxonomy in order to explore the tendencies behind the

use of each procedure/name type;

This objective was met in Chapter 9. Several approaches were taken to present the information and results collected in a clear and concise manner. The chapter is divided into three parts: the first presents the raw and relative frequencies of use of each procedure; the second links the categories in the taxonomy for proper names with the translation procedures. This shows that more changes can clearly be detected with certain types of names than others, for instance Anthroponyms

351 change more than any other type of name. It also highlights the fact that connotative names change more in translation. This fact can be linked to the theories of the translation of proper names explored in Chapter 3; scholars in favour of the translation of proper names, such as Martinet (1982) or Cuéllar Lázaro (2000) also stress the importance of taking into account the connotations and the role of each individual name when attempting to transfer its meaning across to a different language.

The third part takes a more qualitative approach to describing individual uses of certain names in context. This contributes to showing the special role that names take in these novels and how they are used for their connotations. It also shows, once again, the importance of considering each individual referent for each name in its context. As illustrated with the examples in context used throughout this thesis, names are often used in Chick Lit novels to describe special traits, personalities or looks (for instance in sentence like ‘she looked at me with Alice Cooper eyes’ or ‘I feel like Jeremy Paxman today’). This objective materialises on the list of findings, both concrete and general, set out at the end of Chapter 9.

7. To infer implications from the findings and draw

generalizations from them.

These findings link directly with DTS and more concretely with Toury’s initial norm

(1995). The concepts of adequacy and acceptability were used to set the trends followed by the different translators in the TTs examined. The relationships between

352 each segment pair in SL and TL was explored through the analysis of the procedures employed to translate them. All the findings are discussed in Chapter 10 in light of

DTS as well as the other theories exposed in the theoretical parts of this thesis.

11.4 Research questions

In this section I will attempt to address each research question posed in Section 4 of the Introduction with the aim of determining how far they have been answered.

1. to what extent can the cultural background and sense of place

evoked for native readers by a proper name be translated into different

target cultures, with different perceptions, traditions and values?

This research questions really contains two major aspects, on the one hand it is presenting proper names as connotative items and in the other hand it considers the possibility of transferring these connotations into a different culture through translation.

The answer to the first aspect of this research question lies in the genre chosen for the exploration of the subject, Chick Lit. The list of names extracted from the novels selected, and more descriptively the analysis of some of these names in their context undertaken in section 9.4, emphasizes the role that proper names take. Thus, this research is successful to make proper names transpire as definitors of culture used mainly for their connotations with the intention to trigger certain connections in the

353 minds of the intended readers, familiar with the referents. This use of proper names perhaps differs from that more traditional view of proper names as referential elements used solely for a deictic function to specify the specific entity from all of the other entities in its group (that is ‘the Princess Louise pub in Holborn’ states clearly a firm location as opposed to simply the more general ‘a pub’, which refers to no specific entity).

With regards to the transference of the connotations into different cultures, the analysis of the translations of these names showed the different positions taken by the translators at different times. Pointing towards a combination between changing some names and maintaining others as a way to balance the foreignness of the genre with the explicitation of the connotations intended for SL readers. As discussed in Chapter 10, Mayoral (1999/2000) talks about the equilibrium that translators have to find between what is translated and what is explained. All these points perhaps indicate that the more adequate answer to this subquestion is ‘to a certain extent’.

This research emphasizes the different factors at play, both in the creation and the translation of a proper name and comes to stress the role of the translator as a cultural mediator, attempting to breach the gap not only between languages but primarily, and more difficulty, between cultures. The key lies in the search for equilibrium, for maintaining the flavour of the original and transferring across an understandable novel (and an enjoyable and commercially successful one, of course).

354

This research shows the number of ‘constraining factors’ (Munday 2012) at play, already in an original work of literature, and double in a translation. For example, the Chick Lit novels are often taken to represent and describe modern society, in a similar way to what Jane Austen did to early nineteenth century society. These novels are clearly based in the time where they have been created and are set to operate. In a practical way, this means, that all these proper names can potentially represent a ‘cultural bump’ (Leppihalme, 1997) also for native readers from a different time-span. After all, who can guarantee that a middle aged woman in 2222 will remember Adam Ant, for instance?

The value of this research lies precisely in the existence of the question, highlighting the importance of proper names as connotative and cultural elements and the different aspects that need to be taken into consideration for their translation as much as for their study. Taking into consideration the scope of the development of this research question in this project, effort was put to highlight and represent the way in which proper names are used for their connotations. This specific point differentiates this study of proper names from what has been attempted before, where proper names had been considered as single elements. Although the answer to the second subquestion is to some extent more vague, the project successfully traces the work of the translators in their attempts to transfer connotations setting clear tendencies and behaviours for certain types of names (9.4, 9.5 & 9.6).

355

2. how important are external factors such as the context and the

situation when it comes to understanding the connotations conveyed

by a proper name and to what extent is it possible to map these

different levels of information through a classification of proper names

and/or of translation procedures?

The answer to this research question is inextricably linked to that of research question 1. This research primarily shows the importance of the individual referent behind each name when it comes to understanding the full extent of its meaning

(when proper names are used primarily for their connotations, as explained in question 1). This is true both for original names from SL texts and for the translated names. Along the examples used through this study, we have seen a widespread use of proper names, names used mainly to name or identify a character, names to describe a trait or attribute of a person, names to evoke more specific connotations or even to describe a feeling or a situation. With this in mind, the context, the intention and the situation immediately stand out as essential elements to take into account when considering proper names.

If the subquestion hidden behind research question 2 is unpacked: “up to what extent is it possible to map these levels of information through a classification of proper names and/or of translation procedures?” a more concrete answer can be found on this project, and this represents one of the biggest achievements of the thesis. The mapping of the different levels of information exposed in research question 1 was achieved through the use of novel techniques for classifying both

356 proper names types and translation procedures. And the combination of both taxonomies provides a clear picture of what happens is translation with these different levels of information.

This research attempted to model these levels of information through the adoption of a faceted classification for proper names, as opposed to using the more rigid model that has traditionally been used. The different facets in the taxonomy each collects a special aspect at play within each single proper name and it is in the combination of facets where the whole extent of the classification can best be appreciated.

By the same token, through the mapping of the procedures used to translate these names it is possible to understand to what extent all these factors have been transferred across, or more specifically, what do translators do as an attempt to transferring them. The answer again goes around a combination of different elements, as fully explained in Chapters 9 and 10.

Combining both taxonomy and map of procedures it is possible to illustrate what is hidden behind each individual name, as well as the different approaches taken to translate them.

3. to what extent is it possible to infer translation procedures and

understand the processes which lead from source to TTs through

the analysis of translated proper names?

357

As already pointed out by Hermans (1988), proper names are very valuable objects of study for translational norms governing the act of translation. This research again demonstrates this point. The map of translation procedures used in this research was created from an amalgamation of previous maps, but it was ultimately the proper names extracted from the corpus which shaped and finalized the map. It was a circular process of a constant feed of information. Each new book added to the corpus was classified and each new classification fed into the map.

Ultimately, a tendency to conserve certain elements together with changes to some others was revealed and it started to be possible to set the work of Spanish translators mainly towards adequacy and the work of Italian translators mainly towards acceptability. Thus, it can be affirmed that proper names have the potential to contribute towards furthering the understanding of the processes of translation and revealed certain tendencies and patterns of translation.

4. What can a corpus-based approach offer to the study of

translated proper names and to what extent is it beneficial?

Primarily and most importantly, a corpus-based approach offers accuracy and consistency, ensuring a less error-prone account and analysis of the translation procedures used with proper names. Having to locate all the proper names in all nine books included in BRIDGET manually turning the pages of each novel would have been an arduous task probably with lower accuracy overall. By the same token,

358 manually classifying the names and counting and adding up each time a category or a procedure is used would have resulted in more inconsistent figures with several errors along the way. From this point of view, a corpus-based approach is certainly the only approach. The concordancer was the most valuable tool employed in this research. The amount of times that I searched for each name is innumerable, and the benefits of replicating the results each time a search was performed are invaluable.

Meanwhile, this research can also be taken to contribute to the corpus-based approach as the results mimic those conclusions more commonly taking shape in the latest and more ground-breaking projects of the discipline, calling for an inclusion of external factors to the linguistic approaches of the corpus tools and investigations, as pointed out in Chapter 10.

On the other hand, however, the problems with the technology cannot be completely overlooked and highlight perhaps that the discipline is still in its infancy.

Although major advances in computing are contributing to the fast development of tools, there is constant room for more changes and improvements. Thus, again unpacking the subquestion hidden in research question 4 “to what extent is it [the use of a corpus approach] beneficial” it can be said that it is the existence of research projects like this one which makes the discipline go forward contributing new evidence indicating where more work is needed as well as contributing new programs such as Access for BRIDGET. Therefore, making the project beneficial for

359 the discipline. As seen above, in turn, a corpus-based approach, made this project feasible and more comprehensive that it could have been if done manually.

11.5 Limitations and further work

Within the scope and time span of a PhD it is inevitable that there will be limitations to the research. This section will point out some of the limitations to which this study is subject. Furthermore, I will also attempt to link these limitations to future work which could be undertaken to mitigate them.

The first limitation relates to the size of the corpus. Although three novels (9 books in total counting STs and TTs) provided an important amount of distinct proper names and translation instances for the analysis, it is clear that a 653,000 word corpus cannot be comprehensive.

In addition, three novels also have the limitation of a limited number of authors and translators, and similarly a limited number of editors. The possibility of some of these findings being attributable to translator’s or editor’s own preference has to be recognised. More research in the area will mitigate the impact of these limitations.

The initial list of Chick Lit books compiled as the starting point of this research

(Chapter 6) included 22 books (Appendix 1). It was, of course, impossible to include all these books in BRIDGET for two main reasons: firstly, it is practically impossible for a single researcher to scan, proofread and align so many books within the time

360 span of a PhD research project. Secondly, the difficulties with obtaining copyright permission to manipulate the books also affected the final number of books to be included. Subsequently, a first indication for further work points to enlarging

BRIDGET with more books from this list. New books will offer new data which could be contrasted with the results obtained with these three books. More books could be used, for example, to shed light on some of the findings more partially indicated by this study, for instance the different approach which seemed to occur in My Life on a

Plate compared to the other two books in the corpus; was this due to any particular trait of this one book or did the reasons lie more with the particular translators, for instance?

The second limitation is closely related to the first one, as it concerns the fact that

BRIDGET is a unidirectional parallel corpus. Although this point did not affect the research results, as there was a certain degree of variety in the publication time span, authors, translators and book plots, and the corpus was deemed fit for the purpose and aims of this research. Another dimension to the study of proper names in Chick Lit could be given by including in BRIDGET Chick Lit novels written originally in Spanish by Spanish writers (Chapter 5) (for example Miriam Lavilla) as well as in

Italian by Italian writers (Tiziana Mirani, for instance). These could act as a comparable corpus, and could be used to shed new light towards the use of proper names. It is clear from this thesis that proper names play a very special role in

British Chick Lit, but is this the case with Chick Lit originating in other countries?

361

The third limitation covers the object of study of this research, proper names. This research shows that authors of Chick Lit novels use proper names to evoke certain connotations to their readers, and the study of the proper names used in these three novels already pointed out tendencies and patterns of translation employed in these novels. However, cannot be ruled out that other elements may also contain important information and BRIDGET could be used to investigate other linguistic, cultural and translational factors and hypothesis, for instance an analysis of the word-frequency lists could be performed with an aim to locating the most typical language used in these novels.

Following on from these limitations, it is clear, therefore, that areas for further work mainly include the benefit of supplementing BRIDGET with new material, both more books in the same line of the ones already included, as well as comparable ones from the TL countries. In a similar way, of course, adding new languages and translations to BRIDGET to expand the scope of this research can also be another area for further work.

Concerning the same area of study, investigating more specialized features of the

Chick Lit genre and applying more advanced computerized techniques can also be suggestions for further work. The experiments conducted with the means for automatic identification and extraction of proper names from corpus can be taken further and Access for BRIDGET could also be upgraded to look for alternative features and elements in texts.

362

Another area that has been opened up by this research, and has already been pointed out in the limitations section, is the comparison of original British Chick Lit to

Chick Lit from other countries. This could be done with a view to determining if there are any particular features to British Chick Lit language or as measure of the extent of the influence of the British culture in the genre as a whole, especially considering that the genre it is believed to have originated in the UK and has only recently began to produce original pieces in non-English speaking countries.

11.6 Final remarks

In this chapter I have attempted to summarize the main procedures and findings of this research. Despite the inevitable limitations, the objectives stated in the introduction have been met and the research questions have been answered, at least to some extent. Most importantly, the generalizations generated from this analysis of proper names contribute to the corpus of DTS case studies, furthering the discipline, while the corpus and program created are available to be taken forward in future research projects. In addition, the models, maps and processes devised for the analysis, and successfully applied in this research, represent the large scale contribution to knowledge of this investigation and a theoretical advance that will now benefit from replication in other investigations. The taxonomy for proper names can be used in other social science and humanities investigations, therefore expanding the scope of the results of this study to other disciplines and subject fields outside Translation Studies.

363

Bibliography

Primary texts

COLGAN, J. (1999) Amanda’s Wedding, London, HarperCollins.

COLGAN, J. (2003) La boda de Amanda, translated by Chávez, MJ & Figueroa, A., Madrid,

Suma de Letras.

COLGAN, J. (2001) Non sposate quella donna!, translated by Daniele, V., Milan, Salani

Editore.

KNIGHT, I. (2000) My Life on a Plate, London, Penguin.

KNIGHT, I. (2001) Mi vida es casi perfecta, translated by Murillo E.G, Barcelona, Plaza y

Janés.

KNIGHT, I. (2001) La mia vita su un piatto, translated by Dallaglio, A., Milan, Feltrinelli

Editore.

FIELDING, H. (1996) Bridget Jones’s Diary, London, Pan Macmillan.

FIELDING, H. (1998) El diario de Bridget Jones, translated by Busquets, N., Barcelona,

Lumen.

FIELDING, H. (1998) Il diario di Bridget Jones, translated by Crosio, O., Milan, RCS Libri.

364

Secondary sources

AHMAD, K. (2008) Being in Text and Text in Being: Notes on Representative Texts. IN ANDERMAN , G. & ROGERS, M. (Eds.) Incorporating Corpora. The Linguist and the Translator. Clevedon, Multilingual Matters, 60-94.

AIJMER, K. & ALVSTAD, C. (2005) Introduction. IN AIJMER, K. & ALVSTAD, C. (Eds.) New Tendencies in Translation Studies. Selected papers from a workshop. Goteborg, Acta Universitatis Gothoburgensis.

ALGEO, J. (1973) On Defining the Proper Name, Gainesville, University of Florida Press.

ALLERTON, D. J. (1987) The Linguistic and Sociolinguistic Status of Proper Names. What are They and Who do they Belong to? Journal of Pragmatics, 11, 61-92.

ANDERMAN , G. & ROGERS, M. (2008) The Linguist and the Translator. IN ANDERMAN , G. & ROGERS, M. (Eds.) Incorporating Corpora. The Linguist and the Translators. Clevendon, Multilingual Matters, 5-17.

BAINBRIDGE, B. (2001) Bainbridge denounces chick lit as "froth". Guardian Online, 23 August. http://www.guardian.co.uk/books/2001/aug/23/bookerprize2001.bookerprize [Last Accessed: 29/01/2013].

BAKER, M. (1993) Corpus Linguistics and Translation Studies. Implications and Applications. IN BAKER, M., FRANCIS, G. & TOGNINI-BONELLI, E. (Eds.) Text and Technology. In Honour of John Sinclair. Amsterdam and Philadelphia, John Benjamins, 233-250.

BAKER, M. (1996) Corpus-based Translation Studies: The Challenges That Lie Ahead. IN SOMERS, H. (Ed.) Terminology, LSP and Translation. Studies in language engineering in honour of Juan C. Sager. Amsterdam and Philadelphia, Benjamins, 175-186.

BAKER, M. (2000) Towards a Methodology for Investigating the Style of a Literary Translator. Target, 12(2), 241-266.

BAKER, M. (2004) A Corpus-based View of Similarity and Difference in Translation. International Journal of Corpus Linguistics, 9(2), 167-193.

BARONI, M. & BERNARDINI, S. (2004) BootCaT: Bootstrapping corpora and terms from the Web. LREC 2004.

BAUER, G. (1985) Namenkunden des Deutschen, Bern, Germanistische Leherbuchsammlung Brand 21.

BERNÁRDEZ, E. (1987) El nombre propio: su función y su traducción. IN SABIO, F. A. X. E. (Ed.) Problemas de la traducción. Madrid, Fundación Alfonso X el Sabio, 11-21.

BERNARDINI, S. (2003) Corpora in translation education: An Introduction. IN ZANETTIN, F., BERNARDINI, S., STEWART, D. (Ed.) Corpora in Translator Education. Manchester, St Jerome.

365

BLANCO, P. E. (2009) Diccionario Literario: Chick Lit www.papelenblanco.com/diccionario- literario/diccionario-literario-chick-lit [Last Accessed: 29/01/2013].

BOSSEAUX, C. (2001) A Study of the Translator's Voice and Style in the French Translations of Virginia Woolf's The Waves. CTIS Occasional Papers, 1, 55-75.

BOSSEAUX, C. (2007) How Does it Feel?: Point of View in Translation, Amsterdam, Rodopi.

BOWKER, L. (2002) Working with Specialized Language: A Practical Guide to Using Corpora, London & New York, Routledge.

CANTORA, L. (2006) La traducción de los nombres propios en novelas modernas británicas. Proyecto de investigación presentado para la obtención de la Suficiencia Investigadora. Valladolid, Universidad de Valladolid.

CANTORA, L. (2010) Using Named-Entity recognition systems in the literary domain. IN XIAO, R. (Ed.) International Symposium on Using Corpora in Contrastive and Translation Studies. Edge Hill University, Ormskirk, http://www.lancs.ac.uk/fass/projects/corpus/UCCTS2010Proceedings/.

CARTER, R. (1998) Vocabulary. Applied Linguistic Perspectives, London & New York, Routledge.

CHESTERMAN, A. (2004) Beyond the particular. IN MAURANEN, A. & KUJAMÄKI, P. (Eds.) Translation Universals: Do They Exist? Amsterdam and Philadelphia, John Benjamins, 33-49.

CHESTERMAN, A. (2010) Why study translation universals? HELDA - The Digital Repository of University of Helsinki, http://hdl.handle.net/10138/24319 [Last Accessed: 22/10/2012].

COCKS, G. (1993) An Implementation of the nu-G machine runtime system on the GRID parallel processing system, [Unpublished Master of Science Thesis]. London, University College London.

COCKS, G. (2004) Educational delivery channels for cultural content. IN HEMSLEY, J. CAPPELLINI, V. & STANK, G. (Eds.). EVA 2004 London. Proceedings of a conference held in University College London, 26-30th July. EVA Conferences International.

COLGAN, J. (2001) We know the difference between foie gra and hula hoops, Beryl, but sometimes we just want hula hoops. Guardian, 24 August, 2001. http://www.theguardian.com/books/2001/aug/24/fiction.features11 [Last Accessed: 08, September, 2013].

CUÉLLAR LÁZARO, C. (2000) Dobletes de Traducción y Traductología. Las traducciones al castellano en España de la literatura contemporánea en lengua alemana (1945- 1990). Estudio lingüístico, Valladolid, Universidad de Valladolid, Secretariado de Publicaciones e Intercambio Editorial.

CUÉLLAR LÁZARO, C. (2004) Estado actual de la investigación en Traducción Onomástica. Punto y Coma, 89.

366

http://europa.eu.int/comm/translation/bulletins/puntoycoma/89/pyc898_es.htm [Last Accessed 23/08/2012].

CUNNINGHAM, H. (1999) Information Extraction - A User Guide. Research Memo.

CUNNINGHAM, H., MAYNARD, D., BONTCHEVA, K. & TABLAN, V. (2002) A Framework and Graphical Development Environment for Robust NLP Tools and Applications. Proceedings of the 40th Anniversary Meeting of the Association for Computational Linguistics (ACL'02). Philadelphia.

DASTJERDI, H. V. S. S. (2008) Lost in translation: an intertextual study of personal proper- name allusions. Across Languages and Cultures 9(1), 41-55.

DAVIES, E. (2003) A Goblin or a Dirty Nose? The treatment of Culture-Specific References in Translations of the Harry Potter Books. The Translator, 9(1), 65-100.

DELABASTITA, D. (1996) Introduction. Wordplay and Translation. The Translator, 2(2), 127- 139.

ESPINAL, T. (1989-1991) Sobre la traducción de los nombres propios. Cuadernos de Traducción e Interpretación, 11-12, 73-93.

EZARD, J. (2001) Bainbridge tilts at "chick Lit" cult. Guardian, 24 August. http://guardian.co.uk/uk/2001/aug/24/booksgeneralfiction. [Last Accessed: 29/01/2013].

FAWCETT, T. (2006) An introduction to ROC analysis. Pattern Recognition Letters, 27(8), 861-874.

FERNANDES, L. (2004) Brazilian Practices of Translating Names in Children’s Fantasy Literature: A Corpus-based Study. Unpublished Doctoral Thesis. Florianópolis., Universidade Federal de Santa Catarina.

FERNANDES, L. (2006) Translation of Names in Children’s Fantasy Literature: Bringing the Young Reader into Play. New Voices in Translation Studies, 2, 44-56.

FERNÁNDEZ LEBORANS, M. J. (1999) El nombre propio. IN BOSQUE, I. & DEMONTE, V. (Eds.) Gramática descriptiva de la lengua española. Madrid, Espasa, 77-112.

FERRIS, S. & YOUNG, M. (Eds.) (2006) Chick-Lit: The New Woman's Fiction, New York and Abingdon, Routledge.

FIELDING, H. (2013) Bridget Jones vs. Pride and Prejudice. BBC News - 28 January 2013. http://www.bbc.co.uk/news/entertainment-arts-21204956 [Last accessed 8 August 2013].

FIRTH, J. R. (1968) Linguistics and Translation. IN PALMER, F. R. (Ed.) Selected Papers of J. R. Firth, 1952-1959. London and Harlow, Longmans, 84-95.

FLETCHER, B. (2004) Making the Web more useful as a source for linguistic corpora. IN CONNOR, U. & UPTON, T. (Eds.) Corpus linguistics in North America 2002. Amsterdam, Rodopi.

367

FLORIN, S. (1993) Realia in translation. IN ZLATEVA, P. (Ed.) Translation as Social Action. London, Routledge, 122-128.

FRANCIS, G. (1993) A Corpus-Driven Approach to Grammar. IN BAKER, M., FRANCIS, G. & TOGNINI-BONELLI, E. (Eds.) Text and Technology. In Honour of John Sinclair. Amsterdam and Philadelphia, John Benjamins, 136-156.

FRANCO AIXELÁ, J. (1996) Culture-Specific Items in Translation. IN ROMÁN, A. & VIDAL, C. A. (Eds.) Translation, Power, Subversion. Clevendon, Multilingual Matters.

FRANCO AIXELÁ, J. (1998) La traducción por defecto de los nombres propios (inglés- español): una nueva propuesta basada en el análisis de la realidad Sendebar, 8(9), 33-54.

FRANCO AIXELÁ, J. (2000) La traducción condicionada de los nombres propios, Madrid, Almar.

FREGE, G. (1892) On sense and reference. IN MOORE, A. W. (Ed.) Meaning and Reference. Oxford, Oxford University Press, 23-42.

GARCÍA GARCÍA, O. (2001) La onomástica en la traducción al alemán de Manolito Gafotas IN EXTREMADURA, U. D. (Ed.) Anuario de Estudios Filológicos 24. 153-167.

GARCÍA MARÍN, P. (2010) Los referentes culturales de tipo jurídico en la ficción narrativa: Análisis descriptivo en un corpus de novelas en lengua inglesa y su traducción al español [Unpublished PhD Thesis]. Castellón de la Plana, Universidad de Castellón.

GARCÍA YEBRA, V. (1982) Teoría y práctica de la traducción, Madrid, Gredos.

GARDINER, A. (1954) The Theory of Proper Names. A Controversial essay, Oxford, Oxford University Press.

GILL, R. & HERDIECKERHOFF, E. (2006) Rewriting the romance. Feminist Media Studies, 6(4), 487-504.

GORMLEY, S. (2009) Introduction. Working Papers on the Web, Special Issue on 'Chick Lit', 13(http://extra.shu.ac.uk/wpw/chicklit/gormley.html), [Last Accessed: 29/01/2013].

GRASS, T. (2000) Typologie et Traductibilité des Noms Propres del Állemand Vers Le Francais. TAL, 41(3), 644-669.

GRASS, T. (2002) Quoi! Vous voulez traduire 'Goethe'?: essai sur la traduction des noms propres allemands-français Bern, Peter Lang.

GRASS, T., HUMBLEY, J. & VAXELAIRE, J.-L. (Eds.) (2006) La traduction des noms propres, Montreal, Les Presses de l'Université de Montréal.

GRASS, T., MAUREL, D. & PITON, O. (2002) Description of a Multilingual Database of Proper Names. Actes de PorTal 2002, Lecture Notes in Computer Science 2389, 137-140.

368

GRASS, T., MAUREL, D. & TRAN, M. (2004) Prolexbase: une ontologie pour le traitement multilingue des noms propres IN RITA TEMMERMAN, U. K. (Ed.) The translation of domain specific languages and multilingual terminology management 3. 293-309.

GUTIÉRREZ RODRÍGUEZ, M. M. (2004) The problems of translation of proper names in Harry Potter and The Lord of the Rings. English Studies 25, 123-136.

GUTT, E. A. (1998) Pragmatic Aspects of Translation: Some Relevance-Theory Observations. IN LEO, H. (Ed.) The Pragmatics of Translation. Cleavedon, Multilingual Matters, 41- 53.

GUTT, E. A. (2000) Translation and Relevance: Cognition and Context, Manchester, St Jerome.

HALLIDAY, M. A. K. (1966) Lexis as a linguistic level. IN BAZELL, C. E., CATFORD, J. C., HALLIDAY, M. A. K. & ROBINS, R. H. (Eds.) In Memory of J.R. Firth. London and Harlow, Longmans, 148-162.

HALLIDAY, M. A. K. & HASAN, R. (1991) Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective, Oxford, Oxford University Press.

HAREIDE, L. & HOFLAND, K. (2012) Compiling a Norwegian-Spanish Parallel Corpus. IN OAKES, M. & JI, M. (Eds.) Quantitative Methods in Corpus-Based Translation Studies. Amsterdam and Philadelphia, John Benjamins, 75-113.

HARZEWSKI, S. (2006) Tradition and displacement in the new novel of manners. IN FERRIS, S. & YARDLEY, C. (Eds.) Chick Lit: The New Woman's Fiction. New York and Abignon, Routledge, 29-46.

HARZEWSKI, S. (2011) Chick Lit and Postfeminism, Charlottesville, University of Virginia Press.

HERMANS, T. (1988) On Translation of Proper Names, with Reference to ´De Witte and Max Havelaar´. IN WINTLE, M. (Ed.) Modern Dutch Studies. London, Athlone, 11-24.

HERMANS, T. (1995) Revisiting the classics: Toury's empiricism version one. The Translator, 1(2), 215-23.

HERMANS, T. (1999) Translation in Systems, Manchester, St. Jerome.

HOLMES, J. S. (1988/2000) The name and nature of translation studies. IN VENUTI, L. (Ed.) The Translation Studies Reader. London & New York, Routledge, 172-185.

HOUSE, J. (2011) Using Translation and Parallel Text Corpora to Investigate the Influence of Global English on Textual Norms in Other Languages. IN KRUGER, A., WALLMACH, K. & MUNDAY, J. (Eds.) Corpus-Based Translation Studies: Research and Applications. London and New York, Continuum, 187-210.

JALENIAUSKIENĖ, E. & ČIČELYTĖ, V. (2009) The Strategies for Translating Proper Names in Children’s Literature. STUDIES ABOUT LANGUAGES, 15, 31-42.

JESPERSEN, O. (1924/1964) The Philosophy of Grammar, New York, Norton.

369

JOHANSSON, S. (1998) On the role of corpora in cross-linguistic research. IN JOHANSSON, S. & OKSEFJELL, S. (Eds.) Corpora and Cross-linguistic Research: Theory, Method and Case Studies. Amsterdam and Atlanta, Rodopi, 3-24.

JOHANSSON, S. (2003) Reflections on corpora and their uses in cross-linguistic research. IN ZANETTIN, F., BERNARDINI, S. & STEWART, D. (Eds.) Corpora in Translator Education. Manchester, St Jerome, 135-144.

JONES, E. (2008) Inside the world of chick lit. BBC News, 23 August. 23, August http://news.bbc.co.uk/2/hi/entertainment/1504733.stm [Last Accessed: 29/01/2013].

KALVERKAMPER, H. (1995) Namen im Sprachaustaush: Nameubersetzung. IN EICHLER, E., HILTY, G., LOFFLER, H., STEGER, H. & ZGUSTA, L. (Eds.) Namenforschung. Ein internationales Handbuch zur Onomastick, 2 vols. Berlin & New York, Walter de Gruyter, 1018-1025.

KARAMITROGLU, F. (2000) Towards a Methodology for the Investigation of Norms in Audiovisual Translation, Amsterdam and Atlanta, Rodopi.

KEARNS, J. (1997) Translation Strategies. IN BAKER, M. (Ed.) The Routledge Encyclopaedia of Translation Studies. London and New York, Routledge, 282-285.

KENNEDY, G. (1998) An Introduction to Corpus Linguistics, Amsterdam, Rodopi.

KENNY, D. (2000) Looking for Creativity in a Parallel Corpus. IN OLOHAN, M. (Ed.) Intercultural Faultlines. Research Models in Translation Studies 1. Textual and Cognitive Aspects. Manchester, St. Jerome, 92-104.

KENNY, D. (2001) Lexis and Creativity in Translation, Manchester, St. Jerome.

KILGARRIFF, A. & GREFENSTETTE, G. (2003) Introduction to the special issue on the Web as corpus. Computational Linguistics, 29(3), 333-347.

KLAUDY, K. (1997) Explicitation. IN BAKER, M. (Ed.) Routledge Encyclopaedia of Translation Studies. London and New York, Routledge, 104-108.

KNOWLES, J. (2004) Editorial. Diegesis: Journal of the Association for Research in Popular Fictions (Special Issue on Chick Lit), 8, 3-4.

KRIPKE, S. (1970/1980) Naming and Necessity, Oxford, Blackwell.

KRIPPENDORFF, K. (2004) Reliability in content analysis: Some common misconceptions and recommendations. Human Communication Research 30(3),411-433.

KRIPPENDORFF, K. (2008) Testing the Reliability of Content Analysis Data. What is Involved and Why. IN KRIPPENDORFF, K. & BOCK, M. (Eds.) The Content Analysis Reader, California, Sage, 350-357.

KRUGER, A., WALLMACH, K. & MUNDAY, J. (Eds.) (2011) Corpus-Based Translation Studies: Research and Applications, London and New York, Continuum.

370

LAVIOSA-BRAITHWAITE, S. (1997) Investigating Simplification in an English Comparable Corpus of Newspaper Articles. IN KLAUDY, K. & KOHN, J. (Eds.) Transferre Necesse Est. Proceedings of the 2nd International Conference on Current Trends in Studies of Translation and Interpreting. Budapest, Hungary, Scholastica.

LAVIOSA, S. (Ed.) (1998) The Corpus-based approach, Special issue of Meta 43.4.

LAVIOSA, S. (2002) Corpus-based Translation Studies: Theory, findings, applications, Amsterdam, Rodopi.

LAVIOSA, S. (2011) Corpus-Based Translation Studies: Where Does it Come From? Where Is It Going? IN KRUGER, A., WALLMACH, K. & MUNDAY, J. (Eds.) Corpus-Based Translation Studies: Research and Applications. London and New York, Continuum, 13-32.

LEECH, G. (1981) Semantics: the Study of Meaning, Harmondsworth, Penguin.

LEFEVERE, A. & BASSNETT, S. (1990) Introduction: Proust’s Grandmother and the Thousand and One Nights: The 'Cultural Turn' in Translation Studies. IN BASSNETT, S. & LEFEVERE, A. (Eds.) Translation, History and Culture. London & New York, Pinter Publishers, 1-13.

LEPPIHALME, R. (1997) Culture Bumps. An Empirical Approach to the Translation of Allusions, Clevedon, Multilingual Matters Ltd.

LESSING, D. (2001) See "Bainbridge", 2001.

LEVY, J. (1965) Will Translation Theory be of Use to Translators? IN ITALIAANDER, R. (Ed.) Übersetzen: Vorträge und Beiträge vom Internationalen Kongreß literarischer Übersetzer in Hamburg. Frankfurt, Athenäum, 77-82.

LIETZ, G. (1992) Eigennamen in der norwegischen Gegenwartssprache. Probleme ihrer Wiedergabe im Deutschen am Beispiel belletristicher Texte, Frankfurt & Berlin, Lang.

LOIACONO, R. (2011) Il trattamento dei nomi propri nella traduzione di documenti giuridici tra l’italiano e l’inglese. inTRAlinea Special Issue: Specialised Translation II.

MAHER, J. (2007) The post-feminist mystique ('Chick Lit, The New Woman's Fiction' by Suzanna Feriss and Mallory Young, 'Athena's Daughters, Television's New Women Warriors' by Frances Early and Kathleen Kennedy, and 'Action Chicks, New Images of Women in Popular Culture' by Sherre Innes). College Literature, 34(3), 193-201.

MANCZAK, W. (1991) La Nature du nom propre: Prolegomenes. Nouvelle Reveu d'Onomastique, 17, 25-28.

MANINI, L. (1996) Meaningful literary names: their forms and functions and their translation. The Translator 2(2), 161 - 178.

MARCO, J. (2004) Les técniques de traducció (dels referentes culturals) retorn per quedar- nos-hi. Quaderns, revista de traducció, 11, 129-149.

371

MARCO, J. (2007) The terminology of translation: Epistemological, conceptual and intercultural problems and their social consequences. Target, 19(2), 255-269.

MARTINET, H. (1982) Le noms propres dans la traduccion litteraire. Meta, 27(4), 392-400.

MAYORAL, R. (1994) La explicitación de la información en la traducción intercultural. IN HURTADO, A. (Ed.) Estudis sobre la traducció. Castellón, Publiacions de la Unviesitat Jaume I, 73-96.

MAYORAL, R. (1999/2000) La traducción de referencias culturales. Sendebar, 10/11, 67-88.

MAZZA, C. (2006) Who's laughing now? A short history of chick lit and the perversion of a genre. IN FERRISS, S. & YOUNG, M. (Eds.) Chick Lit. The New Woman's Fiction. New York and Abingdon, Routledge, 17-28.

MAZZA, C. & DESHELL, J. (Eds.) (1995) Chick-Lit: Post-feminist Fiction, Florida, FC2.

MCENERY, T. & WILSON, A. (1996) Corpus Linguistics, Edinburgh, Edinburgh University Press.

MILL, J. S. (1843) A System of Logic, London, Harrison and Co. Printers.

MILTON, J. & BANDIA, P. (Eds.) (2009) Agents of Translation, Amsterdam and Philadelphia, John Benjamins.

MLYNOWSKI, S. & FARIN, J. (2006) A Girls’s Guide to Writing Chick Lit, Philadelphia, Quirk Books.

MODLESKI, T. (2008) Loving with a Vengance. Mass-Produced Fantasies for Women (2nd edition). New York and London, Routledge.

MOLINA, L. (2006) El otoño del pingüino. Análisis descriptivo de la traducción de los culturemas., Castelló de la Plana, Publicaciones de la Universitat Jaume I.

MOLINA, L. & HURTADO, A. (2002) Translation Techniques Revisited: A Dynamic and Functionalist Approach. Meta, 47(4), 498-512.

MONTORO, R. (2012) Chick Lit: The Stylistics of Cappuccino Fiction, London, Continuum.

MOYA, V. (2000) La traducción de los nombres propios, Madrid, Cátedra.

MUNDAY, J. (1998) A computer-assisted approach to the analysis of translation shifts. IN LAVIOSA, S. (Ed.) L'approche basée sur le corpus/The corpus-based approach, Special Issue of Meta 43(4).

MUNDAY, J. (2012) Introducing Translation Studies. Theories and Applications, 3rd edition, London and New York, Routledge.

MUÑOZ MARTÍN, R. (1995) Lingüística para traducir, Barcelona, Teide.

NEWMARK, P. (1981) Approaches to Translation, Oxford, Pergamo Press.

372

NEWMARK, P. (1988) A Textbook of Translation, New York, Pentrice Hall.

NORD, C. (2003) Proper Names in Translations for Children. Translation for children, 48(2), 182-196.

OAKES, M. & JI, M. (2012) Preface. IN OAKES, M. & JI, M. (Eds.) Quantitative Methods in Corpus-based Translation Studies. Amsterdam & Philadelphia, John Benjamins, 1-2.

OLOHAN, M. (2004) Introducing Corpora in Translation Studies, London, Routledge.

OLOHAN, M. & BAKER, M. (2000) Reporting That in Translated English, Evidence for Subconscious Processes of Explication? Across Languages & Cultures, 1(2), 141-158.

OSGOOD, E., SUCI, G. J. & TANNENBAUM, P. H. (1957) The Measurement of Meaning, Illinois, University of Illinois.

PEARSON, J. (2003) Using Parallel Texts in the Translator Training Environment. IN ZANETTIN, F., BERNARDINI, S. & STEWART, D. (Eds.) Corpora in Translator Education. Manchester, St. Jerome, 15-24.

PEER, W. V. (1989) Quantitative Studies of Literature. A Critique and Outlook. Computers and the Humanities, 23, 301-307.

PENG, G. (2009) Using Rhetorical Structure Theory (RTS) to Describe the Development of Coherence in Interpreting Trainees. Interpreting, 11(2), 216-243.

PENROD, L. K. (2010) Pottering around: Harry Potter in Translation. Translation and Impersonation, 1(3), 19-29.

PYM, A. (1992) Translation and Text Transfer, New York, Peter Lang.

PYM, A. (2008) On Toury's laws of how translators translate. IN PYM, A., SHLESINGER, M. & SIMEONI, D. (Eds.) Beyond Descriptive Translation Studies. Amsterdam and Philadelphia, John Benjamins, 311-328.

RAW, L. (Ed.) (2012) Translation, Adaptation, Transformation, London and New York, Continuum.

RENOUF, A. (2003) WebCorp: Providing a renewable data source for corpus linguists. Language and Computers, 48(1), 39-58.

RUSSELL, B. (1905) On Denoting. Mind, 14, 479-493.

RUSSELL, B. (1912) The Problems of Philosophy, Oxford, Oxford University Press.

RUSSELL, I. (2010) Pink books make me see red. Guardian Online, 28 July. http://guardian.co.uk/books/booksblog/2010/jul/28/quee-of-teen-award-pink-book- covers#start-of-comments [Last Accessed:29/01/2013].

SALDANHA, G. (2009) Principles of corpus linguistics and their application to translation studies research. Tradumática,

373

7(http://webs2002.uab.es/tradumatica/revista/num7/articles/01/art.htm [last accessed August 2013]).

SALDANHA, G. (2011) Style of Translation: The use of foreign words in translations by Margaret Jull Costa and Peter Bush. IN KRUGER, A., WALLMACH, K. & MUNDAY, J. (Eds.) Corpus-Based Translation Studies: Research and Applications. London and New York, Continuum, 237-258.

SANTOYO, J. C. (1988) La traducción de los nombres propios. IN SABIO, F. A. X. E. (Ed.) Problemas de Traducción. Madrid, Fundación Alfonso X el Sabio, 45-50.

SCOTT, M. (1997) WordSmith Tools versión 2.0, Oxford, Oxford University Press

SCOTT, W. (1955/2008) Reliability of content analysis: The case of nominal scale coding. In KRIPPENDORFF, K. & BOCK, M. (Eds.) The Content Analysis Reader, California, Sage, 347-349.

SEARLE, J. R. (1958) Proper Names. Mind, 67(266), 166-173.

SEARLE, J. R. (1967) Proper Names and Descriptions. IN EDWARDS, P. (Ed.) The Encyclopaedia of Philosophy. New York, McMillan.

SEDLEY, D. (2003) Plato's Cratylus, Cambridge, Cambridge University Press.

SÉGUINOT, C. (1988) Pragmatics and the explicitation hypothesis. Traduction, Terminologie, Rédaction, 1(2), 103-13.

SHAROFF, S. (2004) Towards basic categories for describing properties of texts in a corpus. Proceedings of LREC2004. Lisbon, Portugal.

SHAROFF, S. (2006) Creating General-Purpose Corpora Using Automated Search Engine Queries. IN BARONI, M. & BERNARDINI, S. (Eds.) WaCky! Working papers on the Web as Corpus. Bologna, Gedit.

SHAROFF, S., BABYCH, B. & HARTLEY, A. (2009) "Irregragable Answers": Using Comparable Corpora to Retrieve Translation Equivalents. Language Resources and Evaluation, 43, 15-25.

SINCLAIR, J. M. (1966) Beginning the study of lexis. IN BAZELL, C., CATFORD, J. C., HALLIDAY, M. A. K. & ROBINS, R. H. (Eds.) In Memory of J.R. Firth. London and Harlow, Longmans, 410-430.

SORENSEN, H. (1963) The Meaning of Proper Names, Copenhague, Gad.

SPERBER, D. & WILLSON, D. (1995) Relevance: Communication and Cognition, 2nd edition, Oxford, Blackwell.

THOMAS, M. (2009) Localizing Pack Messages: A Framework for Corpus-Based Cross- Cultural Multimodal Analysis. Unpublished PhD Thesis. University of Leeds.

374

TOURY, G. (1978/2004) The nature and role of norms in literary translation. IN VENUTI, L. (Ed.) The Translation Studies Reader, 2nd edition. London and New York, Routledge, 205-218.

TOURY, G. (1995) Descriptive Translation Studies and Beyond, Amsterdam and Philadelphia, John Benjamins.

TOURY, G. (1999) A Handful of Paragraphs on "Translation" and "Norms". IN SCHAFNNER, C. (Ed.) Translation and Norms. Clevendon, Multilingual Matters, 9-31.

TOURY, G. (2012) Descriptive Translation Studies and Beyond, 2nd Edition, Amsterdam and Philadelphia, John Benjamins.

TYMOCZKO, M. (1998) Computerized corpora and the future of translation studies. IN LAVIOSA, S. (Ed.) L'approche basée sur le corpus/The corpus-based approach, Special issue of Meta 43(4). 652-659.

VÁZQUEZ-AYORA, G. (1977) Introducción a la Traductología, Washington, Georgetown University Press.

VENUTI, L. (1995/2008) The Translator's Invisibility: A History of Translation, London and New York, Routledge.

VENUTI, L. (1997) Unequal Developments: Current Trends in Translation Studies. Comparative Literature Studies, 49(4), 360-368.

VENUTI, L. (Ed.) (2000/2004) The Translation Studies Reader, London and New York, Routledge.

VENUTI, L. (2008) The Translator's Invisibility, London and New York, Routledge.

VIEZZI, M. (2004) Denominazioni proprie e traduzione, Milan, LED Edizioni Universitarie.

VINAY, J.P. & DARBELNET, J. (1958/1995) A Methodology for Translation. IN LAWRENCE, V. (Ed.) The Translation Studies Reader. New York and London, Routledge, 128-137.

WELLS, J. (2006) Mothers of Chick Lit? Women Writers, Readers and Literary History. IN FERRIS, S. & YOUNG, M. (Eds.) Chick Lit: The New Women's Fiction. New York and Abignon, Routledge, 47-70.

WILLEMS, K. & MUSSCHE, E. (2010) Fred or farīd, bacon or bayḍun (‘egg’)? Proper names and cultural-specific items in the Arabic translation of Harry Potter. Meta, 55(3), 474- 498. WILSON, C. A. (2012) Becky Bloomwood at the V&A: Culture, Materialism, and the Chick Lit Novel. Journal of Popular Culture, 45(1), 214-225.

WILSON, J., HARTLEY, A., SHAROFF, S. & STEPHENSON, P. (2010) Advanced Corpus Solutions for Humanities Researchers. Proceedings of the Workshop on Advanced Corpus Solutions, Pacific Asia Conference on Language, Information and Computation PACLIC24. Sendai, Japan, November 2010.

375

YARDLEY, C. (2006) Will Write for Shoes. How to Write a Chick Lit Novel, New York, St. Martin's Press.

ZABEEH, F. (1968) What's in a Name? An Inquiry into the Semantics and Pragmatics of Proper Names, La Haye, Martinus Nijhoff.

ZANETTIN, F. (2000) Parallel Corpora in Translation Studies: Issues in Corpus Design and Analysis. IN OLOHAN, M. (Ed.) Intercultural Faultlines. Research Models in Translation Studies 1. Textual and Cognitive Aspects. Manchester, St. Jerome, 105- 118.

376